Online Program Schedule

The program schedule is subject to change. Check this site for updates. When you arrive at the meeting site, check the final schedule for any last-minute changes.

Session Schedule & Abstracts

Please note that we’re in the process of correcting typographical errors. If you see such errors, please report them to Larry Witmer (, but changes to content will not be made.

Wednesday 29th June, 2016

Morphological Integration & Modularity 1

Room: Salon H   2:30 pm–4:00 pm

Moderator(s): Buchtova M, Gelnaw W
MIM1-1  2:30 pm  Role of FGF signaling during anterior-posterior patterning of zeugopod. Buchtova M*, Laboratory of Molecular Morphogenesis, Institute of Animal Physiology and Genetics AS CR; Institute of Experimental Biology, Faculty of Sciences, Masaryk University, Brno, Czech Republic; Hampl M, Laboratory of Molecular Morphogenesis, Institute of Animal Physiology and Genetics AS CR; Institute of Experimental Biology, Faculty of Sciences, Masaryk University, Brno, Czech Republic; Cela P, Laboratory of Molecular Morphogenesis, Institute of Animal Physiology and Genetics AS CR; Institute of Experimental Biology, Faculty of Sciences, Masaryk University, Brno, Czech Republic; Horakova D, Laboratory of Molecular Morphogenesis, Institute of Animal Physiology and Genetics AS CR, Brno, Czech Republic; Krejci P, Department of Biology, Faculty of Medicine, Masaryk University, Brno, Czech Republic
Abstract: Zeugopod area displays large variability in bone patterning. In some species such as horse, the radius is the main load-bearing bone of the forelimb and the ulna can be significantly reduced in size. On the other hand, in chicken or alligator, the ulna is much larger than the radius. As fibroblast growth factors (FGFs) are key players in processes of cell proliferation and chondrocyte differentiation, we experimentally manipulated this pathway to test its effect on zeugopod modeling. We used loss-of-function approach, where FGFR inhibitor PD173074 was injected into chicken wings or limbs. The inhibitor treatment caused shorter and thinner humerus as well as partial or full absence of radius while ulna remained without morphological changes. The phenotype of hindlimbs resembles the wing phenotype with a more severe effect on the anterior skeleton. The application of PD173074 caused significant changes in cell proliferation and mesenchymal condensation formation during early stages of chondrogenesis in vivo, which was also confirmed in vitro. Moreover, we observed more significant inhibition of chondrogenesis in anterior micromass cultures in comparison to posterior ones. Next, we used gain-of-function approach where recombinant FGF ligands (FGF1 or FGF2) were applied into the right chicken wing bud at stage HH20-22. FGF implantation resulted in skeletal malformations resembling aberrant activation of FGFR/ERK MAP kinase signaling in the growth plate cartilage. Interestingly, humerus and ulna were shorted and thicker, while radius did not exhibit this phenotype. In conclusion, we found that a subtle modification of FGF level affects the size and shape of zeugopod bones during limb development. Based on our results, we propose that similar alteration of FGF signaling could play a role during limb evolution in vertebrates where different degrees of zeugopodial bones reduction are seen. The research was supported by Grant Agency of Czech Republic (14-31540S).

MIM1-2  2:45 pm  Bodies and fins exhibit correlated evolution towards locomotor specializations in cichlid fishes. Feilich K. L.*, Harvard University
Abstract: Cichlids comprise an explosive adaptive radiation of freshwater fishes, with tremendous variation in behavior and morphology. This radiation, and its well-studied evolutionary history, presents a unique opportunity to test hypotheses about how functionally-related traits may exhibit correlated evolution in an appropriate phylogenetic context. In the case of fish locomotion, there are several hypotheses as to how fish bodies and fins may have evolved to support particular locomotor specializations: specifically, one would expect (1) swimming economy specialists to have deep anterior bodies with tapered narrow caudal peduncles, semi-lunate caudal fins, and anteriorly positioned median fins; (2) burst acceleration specialists to have cylindrical stout bodies with broad caudal fins and posteriorly-positioned, high surface area median fins and; (3) maneuvering specialists with laterally compressed, deep body profiles and large median fins with high surface area and broad bases. Using a combination of geometric and linear morphometrics in conjunction with partial-least-squares analysis, it was demonstrated that cichlid bodies, median fins, and caudal fins exhibit correlated shape evolution. Patterns of covariation largely agreed with existing hypotheses of evolution towards locomotor specialization. This is strong evidence that particular fin and body shape combinations have adaptive value. Future work will elucidate how morphology and kinematics vary with each other to produce swimming performance.

MIM1-3  3:00 pm  How many roads lead to Rome? Phenotypic and genetic convergence in two independent lines of mice selectively bred for increases in relative limb bone length. Rolian C*, University of Calgary; Yu S, University of Calgary; Sparrow LM, University of Calgary; Farooq S, University of Calgary; Kucka M, Friedriech Miescher Lab - Max Plack Institute; Beluch WH, Friedriech Miescher Lab - Max Plack Institute; Chan YF, Friedriech Miescher Lab - Max Plack Institute
Abstract: Phenotypic convergence in limb bone morphology is common across terrestrial mammals. Limb proportions likely evolved in parallel in independent lineages under similar selection pressures, for example in the context of locomotor specializations. What is less clear, however, is how many distinct genomic, developmental and phenotypic pathways are possible in the evolution of convergent limb skeleton phenotypes. Put differently, if one were to “play the tape” of limb skeletal evolution again, how often would an ancestral taxon under the same selection pressures show the same changes in a descendant taxon, at phenotypic and genetic levels? Here, we addressed this question by looking at such changes in two independent lines of mice selectively bred for increases in tibia length relative to body mass (the Longshanks mice). Using a combination of genome sequencing, analyses of long bone trabecular architecture, and multivariate morphometric analyses of adult skeletons, we show that, while the two traits under selection (tibia length and body mass) are identical between the two Longshanks lines after 15 generations of selective breeding, changes in other skeletal traits differ between the two lines. Longshanks lines 1 and 2 can be readily distinguished on the basis of skeletal differences in their hands, feet and girdles, as well as in their long bone microarchitecture. These changes are reflected in subtle differences in the underlying genetic architecture, which shows a mix of parallel and lineage-specific genomic changes in regions with known roles in skeletal growth. These results suggest that, even among closely related lineages derived from the same ancestral population, parallel selection for the same skeletal phenotype can cause divergent changes in correlated skeletal traits, most likely through changes in distinct genetic and developmental processes.

MIM1-4  3:15 pm  Determinants of body shape and co-variation among elements of the bony torso in anthropoids (Primates: Anthropoidea). Ward CV*, University of Missouri; Middleton ER, University of Missouri
Abstract: A cornerstone assumption for reconstructing anthropoid evolution is that the hominoid torso differs from that of monkeys, with a broader rib cage and longer, broader pelvis to effect more laterally oriented glenohumeral joints, reflecting adaptation to below-branch arboreality. Furthermore, the transition to the hominin body plan is most commonly thought to have had African ape-like beginnings – long pelvis, short, stiff lumbar spine, cone-shaped rib cage. However, growing evidence for homoplasy in hominoid postcranial form appears to challenge these hypotheses. Further, the fossil record tends to leave only isolated bones, forcing paleontologists to rely on hypothesized co-variation among bones to infer torso structure. We test these assumptions about determinants of torso shape in anthropoids to provide a basis for inferring the evolution of torso form in the fossil record. We combine data on intact torso shape gathered from reconstructed CT scans of 68 extant anthropoids with linear and landmark data from 225 skeletons including ribs, vertebrae, sternum, and pelvis. Data reveal more and different patterns of variation among elements of the torso than previously appreciated, and do not support the idea that reorientation of the shoulder led to uniform modification of the torso. Iliac morphology is not tightly coupled with rib cage shape but is broadly related to lumbar form. Spinal musculature is a stronger determinant of pelvic form than previously recognized. Upper and lower portions of the rib cage are not tightly integrated and appear to respond to different adaptive constraints. These results reflect differing selective pressures across different regions of the torso, and provide data with which to more accurately infer aspects of torso form, and locomotor adaptation, in fossil anthropoids, and may provide clues to understanding variation in body shape across mammals. Supported by NSF BCS 0716244 and Leakey Foundation.

MIM1-5  3:30 pm  Building blocks: functional and developmental modularity in the axial skeleton of Felidae (Mammalia). Randau M*, University College London; Goswami A, University College London
Abstract: Geometric morphometric analyses of the axial skeleton in Felidae have shown heterogeneous ecomorphological specialisation in presacral vertebrae shape related to locomotion and prey size. While anterior vertebrae may either have evolved under stronger phylogenetic constraints or are ecologically conservative, posterior vertebrae, specifically in the post-T10 region, show clearer differentiation between ecomorphs. Whereas these results suggest that distinct vertebral sections are under different selection pressures and/or constraints, shape regionalisation in this complex serial structure can also reflect modular organisation of vertebrae, which may direct respondability to selection. Here we have investigated hypotheses of modularity in the axial skeleton of nine living Felidae species spanning the full extant range in body size and ecology. Analyses were performed on a dataset of 1281 vertebrae composed of 19 vertebral types from 70 specimens. We assessed morphological, developmental, and functional modularity hypotheses both across and within vertebrae. Two-block Partial Least Squares analyses, with and without phylogenetic correction, identified five modules in the presacral axial skeleton (atlas – T1, C6 – T2, C7 – T8, T10 – T11, and T12 – L7), whose identities strongly correspond with hypotheses of morpho-functional modularity. Additionally, correlations between the cervical (atlas – C7) and posterior (T12 – L7) modules reflect timing of vertebral ossification. Intravertebral modularity was explored with analyses of relative eigenvalue standard deviation, RV coefficients, and covariance ratios. Our results support the developmental hypothesis of two widespread intravertebral modules (centrum vs. laminae). Exceptions showing different modular patterns are concentrated on regional morphological boundaries, specifically on the axis, C7 and T1, T8, T13, and L6 and L7. Additionally, the axis and vertebrae in the T10 - L7 region show the greatest overall integration.

MIM1-6  3:45 pm  DisintegratoR: a new R package for evaluating phylogenetic trees while accounting for correlations between character states. Gelnaw WB*, University of Texas at Austin
Abstract: One of the assumptions of phylogenetic analysis is that all of the characters being assessed evolve independently of one another. However, morphological characters may be linked due to a shared developmental or epigenetic process, or because states are selected together because of shared functional or ecological pressures. The interdependence of characters is referred to as morphological integration and has been the basis for many arguments against using morphology to construct phylogenetic hypotheses. Morphological integration diminishes the reliability of phylogenetic analyses because it draws support for phylogenetic hypotheses from different characters out of proportion to the information that they actually carry. The most common method used to remove the effect of morphological integration has been for the investigator to identify a suite of correlated character changes, usually associated with a particular ecomorph, and then down-weight or delete those characters to reduce their collective contribution to tree length. DisintegratoR is an R package that I developed to evaluate, among other things, the likelihood of a phylogenetic tree topology given a Brownian motion model of evolution. What sets DisintegratoR apart from other programs is that it also uses the tree structure and the distribution of character states to create a model of correlations between characters, which it then uses to transform the original data to the expected set of states if there was no correlation between characters. The likelihood of the tree is then evaluated using the transformed data. DisintegratoR is a step forward for the phylogenetic analysis of morphological data because it accounts for morphological integration in an objective way that diminishes the risk of introduced for investigator bias.

[back to schedule]