Studying morphological integration and modularity at multiple levels: concepts and analysis

Although most studies on integration and modularity have focused on variation among individuals within populations or species, this is not the only level of variation for which integration and modularity exist. Multiple levels of biological variation originate from distinct sources: genetic variation, phenotypic plasticity resulting from environmental heterogeneity, fluctuating asymmetry from random developmental variation and, at the interpopulation or interspecific levels, evolutionary change. The processes that produce variation at all these levels can impart integration or modularity on the covariance structure among morphological traits. In turn, studies of the patterns of integration and modularity can inform about the underlying processes. In particular, the methods of geometric morphometrics offer many advantages for such studies because they can characterize the patterns of morphological variation in great detail and maintain the anatomical context of the structures under study. This paper reviews biological concepts and analytical methods for characterizing patterns of variation and for comparing across levels. Because research comparing patterns across level has only just begun, there are relatively few results, generalizations are difficult and many biological and statistical questions remain unanswered. Nevertheless, it is clear that research using this approach can take advantage of an abundance of new possibilities that are so far largely unexplored.     

Hierarchical nature of morphological integration and modularity in the human posterior face

Morphological integration and modularity are important points of intersection between evolution and the development of organismal form. Identification and quantification of integration are also of increasing paleoanthropological interest. In this study, the "posterior face," i.e., the mandibular ramus and its integration with the associated midline and lateral basicranium, is analyzed in lateral radiographs of 144 adult humans from three different geographic regions. The null hypothesis of homogenously pervasive morphological integration among "posterior-face" components is tested with Procrustes geometric morphometrics, partial least squares, and singular warps analysis. The results reveal statistically significant differences in integration. Only loose integrative relationships are found between midline and lateral components of the basicranium, which may indicate the presence of at least two different basicranial modules. This modularity can be interpreted in terms of spatiotemporal dissociation in the development of those basicranial structures, and gives support to hypotheses of independent phylogenetic modifications at the lateral and midline basicranium in humans. In addition, morphological integration was statistically significantly stronger between the middle cranial fossa and the mandibular ramus than between the ramus and the midline cranial base. This finding confirms previous hypotheses of a "petroso-mandibular unit," which could be a developmental consequence of well-known phylogenetic modifications in coronal topology of the posterior face and base in hominoid evolution, related to middle cranial fossa expansion. This unit could be involved in later evolutionary tendencies in the hominid craniofacial system.     

A new phylogenetic test for comparing multiple high‐dimensional evolutionary rates suggests interplay of evolutionary rates and modularity in lanternfishes (Myctophiformes; Myctophidae)

The interplay between evolutionary rates and modularity influences the evolution of organismal body plans by both promoting and constraining the magnitude and direction of trait response to ecological conditions. However, few studies have examined whether the best‐fit hypothesis of modularity is the same as the shape subset with the greatest difference in evolutionary rate. Here, we develop a new phylogenetic comparative method for comparing evolutionary rates among high‐dimensional traits, and apply this method to analyze body shape evolution in bioluminescent lanternfishes. We frame the study of evolutionary rates and modularity through analysis of three hypotheses derived from the literature on fish development, biomechanics, and bioluminescent communication. We show that a development‐informed partitioning of shape exhibits the greatest evolutionary rate differences among modules, but that a hydrodynamically informed partitioning is the best‐fit modularity hypothesis. Furthermore, we show that bioluminescent lateral photophores evolve at a similar rate as, and are strongly integrated with, body shape in lanternfishes. These results suggest that overlapping life‐history constraints on development and movement define axes of body shape evolution in lanternfishes, and that the positions of their lateral photophore complexes are likely a passive outcome of the interaction of these ecological pressures.     

Pectoral girdle and forelimb variation in extant Crocodylia: the coracoid–humerus pair as an evolutionary module

To date, all statements about evolutionary morphological transformation in Crocodylia have essentially been based on qualitative observations. In the present study, we assessed the morphological variation and covariation (integration) between the scapula, coracoid, humerus, radius, and ulna of 15 species of Crocodylidae, Alligatoridae, and Gavialis + Tomistoma using three‐dimensional geometric morphometrics. The results obtained reveal that the variation of elements within species (intraspecific) is large. However, despite this variability, variation across species (interspecific) is mainly concentrated in two dimensions where the disparity is constrained: ‘robusticity’ and ‘twist’ (forelimbs) and ‘robusticity’ and ‘flexion’ (pectoral girdle). Robusticity (first dimension of variation) embodies a set of correlated geometrical features such as the broadening of the girdle heads and blades, or the enlargement of proximal and distal bone ends. The twist is related to the proximal and/or distal epiphyses in the forelimb elements, and flexion of the scapula and coracoid blades comprises the second dimension of variation. In all crocodylians, forelimb integration is characterized by the strong correlations of a humerus–ulna–radius triad and by a radius–ulna pair, thus forming a tight forelimb module. Unexpectedly, we found that the humerus and coracoid form the most integrated pair, whereas the scapula is a more variable and relatively independent element. The integration pattern of the humerus–coracoid pair distinguishes a relatively robust configuration in alligatorids from that of the remainder groups. The patterns of variation and integration shared by all the analyzed species have been interpreted as an inherited factor, suggesting that developmental and functional requirements would have interacted in the acquisition of a semi‐aquatic and versatile locomotion at the Crocodylia node at least 65 Mya. Our findings highlight the need to incorporate the humerus–coracoid pair in biodynamic and biomechanical studies. © 2012 The Linnean Society of London     

On the comparison of the strength of morphological integration across morphometric datasets

Evolutionary morphologists frequently wish to understand the extent to which organisms are integrated, and whether the strength of morphological integration among subsets of phenotypic variables differ among taxa or other groups. However, comparisons of the strength of integration across datasets are difficult, in part because the summary measures that characterize these patterns (RV coefficient and r_PLS) are dependent both on sample size and on the number of variables. As a solution to this issue, we propose a standardized test statistic (a z‐score) for measuring the degree of morphological integration between sets of variables. The approach is based on a partial least squares analysis of trait covariation, and its permutation‐based sampling distribution. Under the null hypothesis of a random association of variables, the method displays a constant expected value and confidence intervals for datasets of differing sample sizes and variable number, thereby providing a consistent measure of integration suitable for comparisons across datasets. A two‐sample test is also proposed to statistically determine whether levels of integration differ between datasets, and an empirical example examining cranial shape integration in Mediterranean wall lizards illustrates its use. Some extensions of the procedure are also discussed.     

Morphological integration and modularity in the hyperkinetic feeding system of aquatic‐foraging snakes

The kinetic skull is a key innovation that allowed snakes to capture, manipulate, and swallow prey exclusively using their heads using the coordinated movement of eight bones. Despite these unique feeding behaviors, patterns of evolutionary integration and modularity within the feeding bones of snakes in a phylogenetic framework have yet to be addressed. Here, we use a dataset of 60 μCT‐scanned skulls and high‐density geometric morphometric methods to address the origin and patterns of variation and integration in the feeding bones of aquatic‐foraging snakes. By comparing alternate superimposition protocols allowing us to analyze the entire kinetic feeding system simultaneously, we find that the feeding bones are highly integrated, driven predominantly by functional selective pressures. The most supported pattern of modularity contains four modules, each associated with distinct functional roles: the mandible, the palatopterygoid arch, the maxilla, and the suspensorium. Further, the morphological disparity of each bone is not linked to its magnitude of integration, indicating that integration within the feeding system does not strongly constrain morphological evolution, and that adequate biomechanical solutions to a wide range of feeding ecologies and behaviors are readily evolvable within the constraint due to integration in the snake feeding system.     

A statistical framework for testing modularity in multidimensional data

Modular variation of multivariate traits results from modular distribution of effects of genetic and epigenetic interactions among those traits. However, statistical methods rarely detect truly modular patterns, possibly because the processes that generate intramodular associations may overlap spatially. Methodologically, this overlap may cause multiple patterns of modularity to be equally consistent with observed covariances. To deal with this indeterminacy, the present study outlines a framework for testing a priori hypotheses of modularity in which putative modules are mathematically represented as multidimensional subspaces embedded in the data. Model expectations are computed by subdividing the data into arrays of variables, and intermodular interactions are represented by overlapping arrays. Covariance structures are thus modeled as the outcome of complex and nonorthogonal intermodular interactions. This approach is demonstrated by analyzing mandibular modularity in nine rodent species. A total of 620 models are fit to each species, and the most strongly supported are heuristically modified to improve their fit. Five modules common to all species are identified, which approximately map to the developmental modules of the mandible. Within species, these modules are embedded within larger "super-modules," suggesting that these conserved modules act as building blocks from which covariation patterns are built.     

THE ONTOGENETIC TRAJECTORY OF THE PHENOTYPIC COVARIANCE MATRIX, WITH EXAMPLES FROM CRANIOFACIAL SHAPE IN RATS AND HUMANS

Many classic quantitative genetic theories assume the covariance structure among adult phenotypic traits to be relatively static during evolution. But the cross-sectional covariance matrix arises from the joint variation of a large range of developmental processes and hence is not constant over the period during which a population of developing organisms is actually exposed to selection. To examine how development shapes the phenotypic covariance structure, we ordinate the age-specific covariance matrices of shape coordinates for craniofacial growth in rats and humans. The metric that we use for this purpose is given by the square root of the summed squared log relative eigenvalues. This is the natural metric on the space of positive-definite symmetric matrices, which we introduce and justify in a biometric context. In both species, the covariance matrices appear to change continually throughout the full period of postnatal development. The resulting ontogenetic trajectories alter their direction at major changes of the developmental programs whereas they are fairly straight in between. Consequently, phenotypic covariance matrices—and thus also response to selection—should be expected to vary both over ontogenetic and phylogenetic time scales as different phenotypes are necessarily produced by different developmental pathways.     

Evolutionary integration of the frog cranium

Evolutionary integration (covariation) of traits has long fascinated biologists because of its potential to elucidate factors that have shaped morphological evolution. Studies of tetrapod crania have identified patterns of evolutionary integration that reflect functional or developmental interactions among traits, but no studies to date have sampled widely across the species-rich lissamphibian order Anura (frogs). Frogs exhibit a vast range of cranial morphologies, life history strategies, and ecologies. Here, using high-density morphometrics we capture cranial morphology for 172 anuran species, sampling every extant family. We quantify the pattern of evolutionary modularity in the frog skull and compare patterns in taxa with different life history modes. Evolutionary changes across the anuran cranium are highly modular, with a well-integrated “suspensorium” involved in feeding. This pattern is strikingly similar to that identified for caecilian and salamander crania, suggesting replication of patterns of evolutionary integration across Lissamphibia. Surprisingly, possession of a feeding larval stage has no notable influence on cranial integration across frogs. However, late-ossifying bones exhibit higher integration than early-ossifying bones. Finally, anuran cranial modules show diverse morphological disparities, supporting the hypothesis that modular variation allows mosaic evolution of the cranium, but we find no consistent relationship between degree of within-module integration and disparity.     

Evolutionary persistence of phenotypic integration: influence of developmental and functional relationships on complex trait evolution

Examination of historical persistence of integration patterns provides an important insight into understanding the origin and evolution of complex traits. Specifically, the distinct effects of developmental and functional integration on the evolution of complex traits are often overlooked. Because patterns of functional integration are commonly shaped by selection exerted by the external environment, whereas patterns of developmental integration can be determined by relatively environment-independent selection for developmental homeostasis, examination of historical persistence of morphological integration patterns among species should reveal the relative importance of current selection in the evolution of complex traits. We compared historical persistence of integration patterns produced by current developmental versus ecological requirements by examining the evolution of complex mandibular structures in nine species of soricid shrews. We found that, irrespective of phylogenetic relatedness of species, patterns of developmental and functional integration were highly concordant, suggesting that strong selection for developmental homeostasis favors concordant channeling of both internal and external variation. Overall, our results suggest that divergence in mandible shape among species closely follows variation in functional demands and ecological requirements regardless of phylogenetic relatedness among species.     

Differing evolutionary patterns underlie convergence on elongate morphology in endemic fishes of Lake Waccamaw, North Carolina

An extensive body of research has recently demonstrated patterns of parallel and/or convergent evolution that arise from divergent natural selection pressures exerted across environmental gradients. These studies, although providing some of our best empirical evidence for natural selection, have focused on rather narrow phylogenetic scopes, more often than not comparing patterns of morphological change among closely‐related taxa within a single genus. Organisms in replicated populations in these studies are often assumed to have accomplished convergence via similar underlying processes. However, such assumptions cannot be made when looking at evolution across broader phylogenetic and ecological spectra. In the present study, we assessed morphological change across a much broader scale to test whether similar evolutionary and developmental patterns underlie convergence. Specifically, we studied morphological change that has occurred in a novel lake environment (Lake Waccamaw, North Carolina, USA) where three phylogenetically‐disparate fishes representing different orders have speciated and independently evolved streamlined morphologies relative to their deeper‐bodied progenitors occupying nearby streams and coastal regions. Geometric morphometric analyses revealed that, although the bulk of shape change between environments is similar across taxa, significant species‐specific responses, concordant with differing expectations based on the ecologies of these taxa, were also found. Moreover, allometry analyses indicated that the developmental patterns underlying this change also differ across taxa. The present study provides evidence that, within a common environment, convergence can be achieved by different evolutionary and developmental patterns in phylogenetically‐ and ecologically‐disparate taxa. Finally, these results contradict the commonly‐held hypothesis that fishes should be more streamlined in streams than lakes and emphasize the need to also consider other environmental characteristics, such as water clarity and physical complexity, in studies of divergence. © 2009 The Linnean Society of London, Biological Journal of the Linnean Society, 2009, 98 , 636–645.     

Phenotypic integration in feliform carnivores: Covariation patterns and disparity in hypercarnivores versus generalists

The skeleton is a complex arrangement of anatomical structures that covary to various degrees depending on both intrinsic and extrinsic factors. Among the Feliformia, many species are characterized by predator lifestyles providing a unique opportunity to investigate the impact of highly specialized hypercarnivorous diet on phenotypic integration and shape diversity. To do so, we compared the shape of the skull, mandible, humerus, and femur of species in relation to their feeding strategies (hypercarnivorous vs. generalist species) and prey preference (predators of small vs. large prey) using three-dimensional geometric morphometric techniques. Our results highlight different degrees of morphological integration in the Feliformia depending on the functional implication of the anatomical structure, with an overall higher covariation of structures in hypercarnivorous species. The skull and the forelimb are not integrated in generalist species, whereas they are integrated in hypercarnivores. These results can potentially be explained by the different feeding strategies of these species. Contrary to our expectations, hypercarnivores display a higher disparity for the skull than generalist species. This is probably due to the fact that a specialization toward high-meat diet could be achieved through various phenotypes. Finally, humeri and femora display shape variations depending on relative prey size preference. Large species feeding on large prey tend to have robust long bones due to higher biomechanical constraints.     

Evaluating evolutionary pressures and phylogenetic signal in earthworms: a case study – the number of typhlosole lamellae in Hormogastridae (Annelida,Oligochaeta)

Rarely have phylogenetic comparative methods been used to study the correlation between phenotypic traits and environmental variables in invertebrates. With the widespread convergence and conservativeness of the morphological characters used in earthworms, these comparative methods could be useful to improve our understanding of their evolution and systematics. One of the most prominent morphological characters in the family Hormogastridae, endemic to Mediterranean areas, is their multilamellar typhlosole, traditionally thought to be an adaptation to soils poor in nutrients. We tested the correlation of body size and soil characteristics with the number of typhlosole lamellae through a phylogenetic generalized least squares (PGLS) analysis. An ultrametric phylogenetic hypothesis was built with a 2580‐bp DNA sequence from 90 populations, used in combination with three morphological and 11 soil variables. The best‐supported model, based on the Akaike information criterion, was obtained by optimizing the parameters lambda (λ), kappa (κ), and delta (δ). The phylogenetic signal was strong for the number of typhlosole lamellae and average body weight, and was lower for soil variables. Increasing body weight appeared to be the main evolutionary pressure behind the increase in the number of typhlosole lamellae, with soil texture and soil richness having a weaker but significant effect. Information on the evolutionary rate of the number of typhlosole lamellae suggested that the early evolution of this character could have strongly shaped its variability, as is found in an adaptive radiation. This work highlights the importance of implementing the phylogenetic comparative method to test evolutionary hypotheses in invertebrate taxa.     

The role of post-natal ontogeny in the evolution of phenotypic diversity in Podarcis lizards

Understanding the role of the developmental pathways in shaping phenotypic diversity allows appreciating in full the processes influencing and constraining morphological change. Podarcis lizards demonstrate extraordinary morphological variability that likely originated in short evolutionary time. Using geometric morphometrics and a broad suite of statistical tests, we explored the role of developmental mechanisms such as growth rate change, ontogenetic divergence/convergence/parallelism as well as morphological expression of heterochronic processes in mediating the formation of their phenotypic diversity during the post-natal ontogeny. We identified hypermorphosis - the prolongation of growth along the same trajectory - as the process responsible for both intersexual and interspecific morphological differentiation. Albeit the common allometric pattern observed in both sexes of any species constrains and canalizes their cephalic scales variation in a fixed portion of the phenotypic space, the extended growth experienced by males and some species allows them to achieve peramorphic morphologies. Conversely, the intrasexual phenotypic diversity is accounted for by non-allometric processes that drive the extensive morphological dispersion throughout their ontogenetic trajectories. This study suggests a model of how simple heterochronic perturbations can produce phenotypic variation, and thus potential for further evolutionary change, even within a strictly constrained developmental pathway.     

Speciation mirrors geomorphology and palaeoclimatic history in African laminate-toothed rats (Muridae: Otomyini) of the Otomys denti and Otomys lacustris species-complexes in the 'Montane Circle' of East Africa

We adopted an integrated systematic approach to delimit evolutionary species and describe phylogeographic, morphometric and ecological relationships in Otomys denti (from the Albertine Rift, Southern Rift in Malawi and the northern Eastern Arc Mountains) and Otomys lacustris (from the Southern Rift in Tanzania and Zambia, and the southern Eastern Arc Mountains). Molecular [cytochrome (cyt) b sequences, 1143 bp, N  = 18], craniometric (classical, N  = 100 and geometric, N  = 60) and ecological (Partial Least Squares regression of shape and ecogeographic variables) approaches show a profound, parallel disjunction between two groups: (1) Eastern Arc and Southern Rift (including the Malawi Rift) ( O. lacustris and Otomys denti sungae ) and (2) Albertine Rift ( Otomys denti denti and Otomys denti kempi ) taxa. Within both groups, cyt b sequences or craniometric analysis provided evidence for the differentiation of both southern and northern Eastern Arc from Southern Rift lineages (across the so-called Makambako Gap). Within the Albertine Rift ( denti – kempi ) lineage, populations from individual mountain ranges differed significantly in skull shape (but not size), but were similar genetically. Over-reliance in the past on very few morphological characters (e.g. number of molar laminae) and a polytypic species concept has obscured phylogenetic relationships and species discrimination in this group. We recognize at least three species in this group, and distinct lineages within two of these species. Each species or lineage was endemic to one of three regions: the Albertine Rift, the Malawi Rift or the Eastern Arc. Our result echo conclusions of recent studies of other mammalian and bird taxa and reflect the geomorphology and palaeoclimatic history of the region.  © 2009 The Linnean Society of London, Biological Journal of the Linnean Society , 2009, 96 , 913–941.     

Postnatal temporal bone ontogeny in Pan,Gorilla, and Homo,and the implications for temporal bone ontogeny in Australopithecus afarensis

Assessments of temporal bone morphology have played an important role in taxonomic and phylogenetic evaluations of fossil taxa, and recent three‐dimensional analyses of this region have supported the utility of the temporal bone for testing taxonomic and phylogenetic hypotheses. But while clinical analyses have examined aspects of temporal bone ontogeny in humans, the ontogeny of the temporal bone in non‐human taxa is less well documented. This study examines ontogenetic allometry of the temporal bone in order to address several research questions related to the pattern and trajectory of temporal bone shape change during ontogeny in the African apes and humans. We further apply these data to a preliminary analysis of temporal bone ontogeny in Australopithecus afarensis. Three‐dimensional landmarks were digitized on an ontogenetic series of specimens of Homo sapiens, Pan troglodytes, Pan paniscus, and Gorilla gorilla. Data were analyzed using geometric morphometric methods, and shape changes throughout ontogeny in relation to size were compared. Results of these analyses indicate that, despite broadly similar patterns, African apes and humans show marked differences in development of the mandibular fossa and tympanic portions of the temporal bone. These findings indicate divergent, rather than parallel, postnatal ontogenetic allometric trajectories for temporal bone shape in these taxa. The pattern of temporal bone shape change with size exhibited by A. afarensis showed some affinities to that of humans, but was most similar to extant African apes, particularly Gorilla. Am J Phys Anthropol 151:630–642, 2013. © 2013 Wiley Periodicals, Inc.     

The impact of digging on craniodental morphology and integration

The relationship between the form and function of the skull has been the subject of a great deal of research, much of which has concentrated on the impact of feeding on skull shape. However, there are a number of other behaviours that can influence craniodental morphology. Previous work has shown that subterranean rodents that use their incisors to dig (chisel‐tooth digging) have a constrained cranial shape, which is probably driven by a necessity to create high bite forces at wide gapes. Chisel‐tooth‐digging rodents also have an upper incisor root that is displaced further back into the cranium compared with other rodents. This study quantified cranial shape and upper incisors of a phylogenetically diverse sample of rodents to determine if chisel‐tooth‐digging rodents differ in craniodental morphology. The study showed that the crania of chisel‐tooth‐digging rodents shared a similar place in morphospace, but a strong phylogenetic signal within the sample meant that this grouping was nonsignificant. It was also found that the curvature of the upper incisor in chisel‐tooth diggers was significantly larger than in other rodents. Interestingly, most subterranean rodents in the sample (both chisel‐tooth and scratch diggers) had upper incisors that were better able to resist bending than those of terrestrial rodents, presumably due to their similar diets of tough plant materials. Finally, the incisor variables and cranial shape were not found to covary consistently in this sample, highlighting the complex relationship between a species’ evolutionary history and functional morphology.     

Mosaic heterochrony and evolutionary modularity: the trilobite genus Zacanthopsis as a case study

Logical connections exist between evolutionary modularity and heterochrony, two unifying and structuring themes in the expanding field of evolutionary developmental biology. The former sees complex phenotypes as being made up of semi-independent units of evolutionary transformation; the latter requires such a modular organization of phenotypes to occur in a localized or mosaic fashion. This conceptual relationship is illustrated here by analyzing the evolutionary changes in the cranidial ontogeny of two related species of Cambrian trilobites. With arguments from comparative developmental genetics and functional morphology, we delineate putative evolutionary modules within the cranidium and examine patterns of evolutionary changes in ontogeny at both global and local scales. Results support a case of mosaic heterochrony, that is, a combination of local heterochronies affecting the different parts individuated in the cranidium, leading to the complex pattern of allometric repatterning observed at the global scale. Through this example, we show that recasting morphological analyses of complex phenotypes with a priori knowledge or hypotheses about their organizational and variational properties can significantly improve our interpretation and understanding of evolutionary changes among related taxa, fossil and extant. Such considerations open avenues to investigate the large-scale dynamics of modularity and its role in phenotypic evolution.