•"Functional morphology" studies the connection between organismal form, behavior, and dynamics.
•Upon scrutiny, the term encompasses at least three distinct approaches: Biomechanics, theoretical morphology, and paleontological applications.
•Morphospace is the theoretical universe of possible organismal morphologies. Depending on one's approach, it can be defined in terms of measured anatomical features or through multivariate statistical analysis.
•Morphospace resembles the "adaptive landscape" metaphor of Sewell Wright.
•Ecological morphospace represents the morphospace occupied by inhabitants of a given environment.
•So much in functional morphology depends on the assumption that natural selection is the sole determinant of morphology, however this "adaptationist program" is regarded as naive by some. Structural constraints, evolutionary heritage, pleiotropy, competing selective pressures, and generic variation can all impart their own non-adaptive signal.
•Testable functional hypothesis can be difficult to develop.
•The comparative method - the interpretation of one organism's features by analogy with those of a better known organism - is a tradition in biology, but can be difficult to implement.
•Some hypotheses are clearly testable, including whether a particular function would require a biological material to exceed its known strength or a muscle to to exceed its known energy output.
"Whether it be the sweeping eagle in his flight, or the open apple-blossom, the toiling workhorse, the blithe swan, the branching oak, the winding stream at its base, the drifting clouds,
over all the coursing sun, form ever follows function, and this is the law. Where function does
not change form does not change."
(Louis H. Sullivan, 1896.)
In general, functional morphology is the analysis of the mechanical and evolutionary relationship of anatomical form to organismal behavior and dynamics. As such the topic is of great interest to paleontology, holding out the hope that a sufficiently rigorous analysis of a fossil taxon's form might enable us to infer its behavior in detail.
Origins and biases: But why would we even expect there to be a connection between form and function? Our faith in this idea springs from three sources:
Natural Selection: The discovery of evolution by natural selection held out a mechanistic alternative, the inexorable effects of natural selection operating over time on every part of the organism which, in principle, should drive it toward an optimum morphology for its environment.
Observation: Quite often, we see organisms with disparate phylogenetic origins converge on the same morphology. E.G.: Sessile Benthic suspension feeders (right).
This only seems explainable in terms of the primacy of natural selection.
However, When examined in detail, people calling themselves functional morphologists don't all do the same things. In fact, three broad categories of related but distinct research tend to get conflated under this term:
Biomechanics: The in vivo study of the structural and dynamic attributes of living creatures, involving direct observations and measurements of living organisms:
Theoretical vs Functional Morphology: The description, in abstract terms, of morphospace - the universe of possible organismal morphologies. Morphospace is, by necessity, an abstraction. Typically, the range of morphological variation in comparable organisms is reduced to two or three principle axes. These may be:
The degree to which these favored measurements truly encompass morphological diversity is not knowable. Raup attempted to reduce coiled mollusk shell morphology to three parameters:
D = distance of aperture from axis of coiling
W = rate of expansion of aperture
T = rate of translation of aperture along axis of coiling
Tweaking these parameters produces a wide range of shapes. But as the Murex example at right shows, these parameters can vary at different points along the aperture and during growth to produce a far greater range of functionally significant forms.
Raup's mollusk adaptive landscape
Morphospace and adaptive landscapes:
Despite its limitations, Raup's initial work revealed a pattern that has been observed many times: The actual distribution of known organisms is often clustered within a specific region of theoretical morphospace. With Raup 1967, the distribution of known planispiral mollusks (ammonoids, primarily) clustered around two peaks (one evolute and one involute planispiral) and entirely avoided the region of morphospace in which successive whorls were not in physical contact. In contrast the spirula, with an internal shell sits well within this otherwise forbidden zone. Invokes two lines of thought:
Functional explanations for the specific observed pattern (structural integrity, drag, prevent external shells from adopting this form whereas an internal shell is not so constrained.)
The adaptive landscape, and much of functional morphology, rests on the notion that adaptation under the influence of natural selection is the paramount determinant of morphology. For much of the 20th century, morphologists counted on this. Could they reasonably do so?
Gould and Lewontine (1979) took aim at the notion, and the reflexive tendency of "pan-selectionist" researchers to offer adaptive secenarios for every aspect of morphology. They compared this to attempts to infer structural function to spandrels, sections of wall filling spaces between load-bearing components like arches and domes in classic Medieval churches. (See the spandrels of San Marco right) In fact, spandrels are just space-fillers between functional elements.
A biological comparison: the human chin - a functionless feature that is a developmental consequence of the reduction of the anterior tooth row.
Structural constraints: By their basic anatomy and margin accretion growth strategy, adult mollusksmust retain the skeleton they had as juveniles, even if they would be better off not having to carry it around.
Evolutionary heritage: Organisms must work with what they inherit. Indeed the variety of ways in which they have addressed biomechanical problems - each in some way suboptimal but all good enough - testifies to that fact. E.G.: rapid swimming in vertebrates and cephalopods. Rapid undulations of the body is much more energetically efficient than cephalopod "jet-propulsion," in which they must expend energy hauling their reaction mass (water) around inside their mantles before expelling it; yet there cephalopods are, the victims of their ancestral heritage, doing the best they can.
Pleiotropy: Some features may be suboptimal because they are coded in the same gene that codes for a highly beneficial trait. Consider aggression and genitalia in hyenas.
Structures may have more than one function: In that case, their form represents a compromise between adaptive trends for each function. Consider wydah tails. The male must use this structure both to attract a mate, and in flight whereas the female only needs to fly with hers.
Limits to genetic diversity: Just because a given change is adaptive doesn't mean that a species has the genetic material with which to effect it.
The result is that few structures are adaptively optimal, as in the classic panda's thumb. (Another structure constrained by evolutionary history.)
Testing functional hypotheses:
So how can we test a functional hypothesis for a morphology? Prothero provides a consensus list of methods:
Define and diagnose the adaptation. This enables you to identify creatures that may have it
Employ phylogenetic information to determine whether the "adaptive feature" is actually a derived evolutionary state.
Consider whether the feature is a "spandrel" - an inevitable artifact of other features.
In a paleontological setting, propose a hypothesis to explain the structure's function. (Once the above issues have been addressed.) This can then be tested using:
Structures that appear functionally analogous are our starting point for the development of hypotheses about life-style. E.G. numerous anatomical characters demonstrate that the extant timber wolf and the extinct native North American dire wolf (right) are closely related and generally similar. They differ mostly in that the dire wolf's skull and teeth were more robust - a little closer to what one sees in spotted hyaenas or other bone-crushing animals.
Thus, we speculate that the dire wolf was generally similar to the timber wolf, but better adapted to bone crushing. Because we are dealing with an organism and with biomechanical functions that are well understood, this contention can be evaluated with any of the methods listed above.
With less familiar creatures, or creatures that are less biomechanically constrained, we face greater difficulties. Consider the Permian bolosaurid reptile Eudibamus cursoris(Berman et al., 2000). Considered, largely on the basis of its small forelimb to hindlimb length ratio to be facultatively bipedal. (Indeed, bipedal lizards typically do show such a ratio.)
Alas, so do other, ecologically dissimilar reptiles, including vertical clingers and leapers like Boyd's forest dragon. In this light, speculations on Eudibamus' locomotion seem reckless without taking a wider range of variables into account.
Doing it better: There are several multivariate statistical analysis methods including principal component analysis that can identify the major sources of variation among specimens. These will be discussed in a later lecture. For now, understand that like the examples given above, they provide a basis for comparison of extinct and living taxa. Unlike them, they also provide an algorithmic approach to identifying and weighting the variables of greatest interest. A recent example: Bennett, 2020 who determined that the enigmatic tiny archosaur Scleromochlus was specialized for leaping similarly to frogs.
What happens when we try to interpret the fossils of an animal with no obvious modern analog?
Consider Parasaurolophus, a Cretaceous lambeosaurine ornithopod dinosaur. Like other ornithopods, it had a deep, laterally compressed torso and tail. The vertebral column of both was stiffened by ossified tendons. The comparison of the deep flat tail to those of swimming vertebrates, combined with the early 20th century conviction that large dinosaurs would have had trouble generating the energy to support their weight on land gave rise to reconstructions of Parasaurolophus as an aquatic creature. The idiosyncratic crest must, therefore, have been a snorkel.
This, at least, was a falsifiable hypothesis. More thoughtful biomechanical analyses of the vertebral column showed that the trunk and tail were inflexible from side to side. Furthermore, the tail vertebrae of actual aquatic reptiles like crocodilians generally had long lateral extensions to give axial muscles better leverage.
Worse, although the crest connected to the pharynx and nasal cavity though an elaborate system of passages, no specimen of the "snorkel" actually had a hole in the top to admit air. Eventually a revised interpretation of Parasaurolophus as a land animal emerged.
Still, its crest stimulates speculation, including.
According to Weishampel, 1997 resonating chambers in the nasal passages facilitated vocalizations. Weishampel modelled these passages in PVC pipe, creating an instrument that produced sound in a low register similar to that of living elephants.
Interesting ideas, all based in some way on the comparative method and absolutely non-falsifiable
Perceiving comparative similarities, like the perception of any pattern in a set of data, is a creative act. Ironically, of all paleontological methods, this one probably has the greatest tendency to push the limits of the proper Scientific Method. In the worst case, the result can be "Just so stories" like those of Rudyard Kipling - appealing speculation with no hope of ever being rigorously tested.
Perhaps the only absolute criterion for falsification of functional hypotheses. A hypothesis of function is definitively falsified if:
It requires a biological material to exceed its known strength. Thus, Farlow et. al, 1995 demonstrated that an adult Tyrannosaurus rex could not withstand a fall from a standing position.
Requires muscles to exceed the known limits of their energy output.
This allows us to make some inferences with confidence. Consider the maximum size of a flying bird: Selective pressure to develop maximum flight thrust is such that in most birds, the major flight muscles already occupy the maximum possible proportion of overall body mass, regardless of size. Thus, they cannot be scaled up allometrically. Huge birds that would require more muscle than this are forbidden. This places a limit on the overall size of an exclusively powered flyer of roughly 12 kg. - roughly the size of the largest powered flyers such as:
are the limit. Any larger bird such as the 15 kg. Pleistocene Teratornis merriami must employ some kind of soaring strategy, at least some of the time. Thus, we are confident that a 40 kg. Pteranodon would also have to resort to soaring because of the same constraints.
Indeed, mapping pterosaurs onto theoretical morphospace (see Witton and Habib, 2010, for example) might support this scenario, provided our morphological assumptions are correct and in agreement.
S. Christopher Bennett. 2020. Reassessment of the Triassic archosauriform Scleromochlus taylori: neither runner nor biped, but hopper. PeerJ 8:e8418.
David S. Berman, Robert R. Reisz, Diane Scott, Amy C. Henrici, Stuart S. Sumida, Thomas Martens. 2000. Early Permian Bipedal Reptile. Science 290,(5493): 969-972.
Paley, William. 1802. Natural Theology; or, Evidences of the Existence and Attributes of the Deity. Collected from the Appearances of Nature.Darwin Online
Imran A. Rahman, James O'Shea, Stephan Lautenschlager, and Samuel Zamora. 2020. Potential evolutionary trade-off between feeding and stability in Cambrian cinctan echinoderms. Paleontology, 30 June 2020.
Adrian L. R. Thomas, Graham K. Taylor, Robert B. Srygley, Robert L. Nudds, Richard J. Bomphrey. 2004. Dragonfly flight: free-flight and tethered flow visualizations reveal a diverse array of unsteady lift-generating mechanisms, controlled primarily via angle of attack. Journal of Experimental Biology 207: 4299-4323.