EVOLUTION AND DEVELOPMENT I: SIZE AND SHAPE

First, some general background to the study of development and evolution. Evolution of organisms involves a change in the developmental program, a change in a series of developmental processes. We often refer to evolution as "descent with modification" and the modification we often notice first is the overall appearance of the organism. This appearance is a result of the development of the organism, thus evolution is intricately involved with development.

Embryology played a major role in evolutionary theory in the 19th century, but was largely ignored in the 20th. Development never really became part of the modern synthesis. Some argue that this is due to the lack of communication between geneticists and developmental biologists. The geneticists were concerned with the rules of transmission of genetic material between generations and the developmentalists were concerned with cellular changes that led to the transformation of an egg into an adult organism. Mutations in adult phenotype were readily available for the study of genetics, but there were precious few "developmental mutants" that bridged the gap between development and genetics (such mutants were discovered in growing numbers during the formulation of the "Modern Synthesis", and many more discovered later).

The general approach is the same as we have taken with the evolution of other traits: development has a genetic basis, if there is genetic variation for the developmental program then development can evolve. We will first take a descriptive approach to evolution and development and next lecture look at some of the more genetic and cellular mechanisms of development.

Early embryologists noticed similarities between ontogeny (the development of an organism) and phylogeny (ancestor descendant relationships in a group). The common phrase "ontogeny recapitulates phylogeny" was put forward by Haeckel as his biogenetic law (see fig. 21.3, pg. 588). Haeckel held that descendants, during their ontogeny, passed through stages that resembled the adults of their ancestors. Before this, Cuvier (1812) held that there were four major classes of organisms: vertebrates, mollusks, articulates and radiates. Cuvier noticed that there was nothing in the ontogeny of a vertebrate that resembled the adult stages of, say, a mollusk. This is because evolution is a bush or a tree not a "ladder" of the great chain of beings. This "branch-like" pattern to phylogeny was apparent to Haeckel, but he still claimed there was "recapitulation".

von Baer made observations about ontogeny and phylogeny that seem obvious to us today, but they are important in development and evolution as they run counter to recapitulation: 1) more general characters appear early in development, 2) less general forms develop from the more general forms, 3) embryos do not pass through other forms they diverge from them, 4) embryos of higher forms only resemble embryos of other forms (human, calf, chick and fish look similar at embryo stage but diverge quickly). See section 17.8.2, and fig. 17.11, pgs. 478-479.

Putting these two views together, we see that there can be a sort of recapitulation within a lineage (i.e., within an evolutionary sequence of ontogenies) but there are many examples that refute the notion that phylogeny is reviewed during ontogeny.

First efforts to place development and evolution in a quantitative, descriptive context were provided by d'Arcy Thompson in On Growth and Form. Using simple rules of geometric transformation he showed that one could obtain the varied forms of organisms by "warping" or "bending" the relative positions of their body parts (see fig. 21.10, pg. 599). These types of diagrams are helpful in identifying what changes of form have taken place, but they do not identify how developmental mechanisms have evolved (the same criticism might be leveled towards Raup's computer snails (see figures 13.7 and 13.8, pgs. 356-357), but mechanism was not the intention of these approaches).

One thing Thompson and Raup's diagrams did contribute was to focus attention on the notion of size and shape. These two very simple words are deceptively complex in the context of the evolution of development. A general paleontological pattern is Cope's rule which states that the body sizes of species in a lineage of organisms tend to get bigger through time. Horse evolution is a classic example. But what happens when you get bigger? In most cases body parts do not grow at the same rate, thus we have allometry.

Allometric growth is the differential rates of growth of two measurable traits of an organism (often it is described as size-correlated changes in shape). It is quantified as y = bxa where x is the measure of one trait, b is a constant, a is the allometric coefficient and y is the other trait. In this form it describes a logarithmic relationship. It can be made into a linear relationship by taking the logs of the values measured for each trait (or by plotting on log x log graph paper):

log y = log b + a log x. This is the equation for a strait line with a being the slope of the line. When a<1 we have negative allometry which means that as x gets bigger, y gets bigger at a smaller rate. When a >1 we have positive allometry which means that as x gets bigger, y gets bigger at a faster rate. When a=1 we have isometry (or isometric growth) which means that there is no change in shape (i.e., the relative sizes of body parts) during growth. See fig. 21.9, pg. 597.

We can describe different kinds of allometry: 1) interspecific allometry where traits of individuals of the same age (usually adults) are compared between different species, 2) intraspecific allometry where a) traits of individuals of all ages are compared within a species (also called ontogenetic allometry), or b) traits of individuals of the same age are compared within a species (also called static allometry).

Some examples: interspecific=the Irish elk example (more below), intraspecific (static)=measurements of body height and arm length in class, intraspecific (ontogenetic)=measurements of body height and arm length with my daughter's day-care measurements included. See figure demonstrating ontogenetic and interspecific allometry of brain and body weight in the same graph.

Intraspecific allometry just describes growth, and alone is not an evolutionary comparison. It is of interest that the allometric coefficient of Bio 48 males and females is ~ 1.0, but if the toddler data are included the allometric coefficient goes up to ~ 1.3. This means that as adults we have about the same proportions (a=1) but as we grow from infant to adult, our arms get proportionally longer (a=1.3).

Allometry is useful in describing the evolution of size and shape. Different species attain different morphologies by virtue of different timing of various developmental processes. This change in timing is called heterochrony. Figures 21.5 - 21.8 and table 21.1, pgs. 590-594 review some of the typical examples of heterochrony. Using the figure below, we can group these into two general classes: in figs B and C the ancestor (dotted) and descendant (solid, but hard to see in C) have the same slope but the descendant stops growing (=adult) at a different time; in figure D and E, the descendant grows for the same amount of time (in these cases same amount of x but different amount of y) but at a different slope. Both are heterochronic changes because some aspect of timing (relative or absolute) has changed in evolution.

Notice that each axis of these graphs include both a measurement component and a time component simply because growth by definition is both a temporal and a dimensional phenomenon. Note that only one of these examples of different growth plans (graph B) demonstrates "ontogeny recapitulates phylogeny": hypermorphosis. Therefore, shape changes can be observed as 1) changes in the slope of an allometric relationship, or 2) changes in the y intercept of an allometric relationship. Recalling high school algebra, a change in the slope will change the intercept, but the intercept can be changed without changing the slope (keep the line parallel and move it up or down). All of these changes result in change in shape. Even with the same slope but different intercepts the relative sizes of x and y will be different so there will be a change in shape. The only case where there is a change in size with no change in shape is when the allometric slope = 1.0 and growth continues (or retards) relative to the ancestor.

Classic examples of allometry are neoteny in human evolution: as adults we look like the juvenile stages of chimps; and neoteny in salamanders: the adult of descendant retains gills (a juvenile morphology in the ancestor). Peramorphosis in the evolution of deer: the "Irish elk" (actually a deer) has phenomenally large antlers and are "disproportionately" large because there is an allometric relationship between body size and antler size. In fact the Irish elk falls right on the line of allometry for other species in the family (interspecific allometry). Thus, the antlers are larger than usual, but they follow precisely the developmental program that seems to be a part of its phylogenetic group. Previous adaptive (and maladaptive) stories had been told about these huge antlers and how they probably drove the elk to extinction, thus a challenge for "adaptive" evolution, but the allometry shows that they are not really "abnormal" (probably went extinct due to climatic changes and hunting). See discussion on pg. 356-358 of adaptive, vs. non-adaptive explanations of morphology.

Allometry is also important in the context of the criticisms to the "adaptationist program". If you looked at a Titanothere with its bizarre horns pointing out of its snout, you might say "what are those things for" as if they evolved for some function. They may not be "for" anything but simply the result of a positive allometric relationship between body size and horn size during evolution. Now the question becomes: what causes Cope's rule? Since allometry is so common, changes in size will produce changes in shape.