American Society of Naturalists

A membership society whose goal is to advance and to diffuse knowledge of organic evolution and other broad biological principles so as to enhance the conceptual unification of the biological sciences.

“Developmental constraints do not influence long-term phenotypic evolution of marsupial forelimbs as revealed by interspecific disparity and integration patterns”

Posted on

Alberto Martín-Serra and Roger B. J. Benson (March 2020)

Phenotypic disparity and integration patterns of marsupial limbs cast doubts on the role of developmental constraints

Read the Article

The developmental constraints of organisms may constrain their ability to evolve new morphologies. Marsupials have low ecological variation and diversity compared to placental mammals, and one influential hypothesis explains this by invoking constraints imposed by their unusual reproductive strategy. In particular, the variability of marsupial arms (forelimbs) might be limited because of their early development and use during the newborn’s arduous crawl to the mother’s pouch. This hypothesis makes several predictions about variation among marsupial species that are analyzed in this study from a macroevolutionary perspective by Alberto Martín-Serra and Roger Benson, using 3D digital models of the limb skeletons of many marsupials. If this hypothesis is correct, then the morphological disparity of marsupial forelimbs should be lower than for their hindlimbs, co-evolutionary integration between fore- and hindlimb should be weaker than within the forelimb, and both patterns should be more conspicuous for diprotodontians (i.e., koalas, wombats, wallabies, kangaroos, and their relatives) because they perform the hardest crawl as newborns. However, the new analyses provide no support for any of these predictions. This indicates that the constrained developmental patterns of marsupial forelimbs have not strongly influenced their long-term evolution history. Therefore, we may have to look for alternative explanations for the differences in ecological and taxonomical diversity between marsupials and placentals, such as geographical distribution and ecological opportunities. Finally, in a broad sense, this work reveals that developmental constraints may be less important than expected when natural selection can act over long timescales.


Marsupials show a smaller range of forelimb ecomorphologies than placental mammals, and it is hypothesized that this results from macroevolutionary constraints imposed by the specialised reproductive biology of marsupials. Specifically, the accelerated development of the marsupial forelimb allows neonates to crawl to the mother’s pouch, but may constrain adult morphology. This hypothesis makes three main predictions: (i) that marsupial forelimbs should show less interspecific disparity than their hindlimbs; (ii) that morphological integration within the marsupial forelimb is stronger than integration between limbs; and (iii) that these patterns should be strongest in diprotodontians, which undergo the most rigorous crawls as neonates. We use a three-dimensional geometric morphometric dataset of limb bones for 51 marsupial species to tests these predictions. We find that (i) marsupial forelimbs and hindlimbs show similar disparities, (ii) no clear differences in integration exist either within or between limbs, and (iii) the same patterns occur in diprotodontians as in other marsupials, even correcting for lineage age. Therefore, there is currently little evidence that the developmental biology of marsupials has constrained their macroevolutionary patterns. Possibly, functional selection can overcome the effects of developmental constraint on macroevolutionary timescales. Our findings suggest that the role of developmental constraints in explaining the limited phenotypic variability of marsupials (compared with placentals) should be reconsidered.