Beyond Heterochrony: The Evolution of Development

Beyond Heterochrony: The Evolution of DevelopmentM. L. Zelditch, (ed.). Beyond Heterochrony: The Evolution of Development. Wiley-Liss, ISBN 0-471-37973-5.

One of the many compensations for spending one's career intimately engaged in a particular academic pursuit is the opportunity to watch intellectual cycles complete themselves. Back in 1977, as I was preparing to enter graduate school with the intention of picking up a quick MSc degree, I was browsing in the bookshop at Southern Methodist University and happened upon a (then) newly published book by one S. J. Gould with the somewhat cryptic title Ontogeny and Phylogeny. While I recognized that these topics were unlikely to impress the folks down at Mobile or Exxon, I bought it anyway because I enjoyed reading about evolutionary theory and had a secret hope of trying to work aspects of evolutionary analysis into my thesis. For me, as for many, that book opened up entirely new ways of looking at the relation between development and evolution. In particular, Gould's "clock model" for representing the relation between size, shape, and developmental age seemed to provide a powerful analytic tool for testing hypotheses designed to clarify the role of development in morphological transformations. Gould's discussion of life-history patterns as potential targets of natural selection was also pregnant with meaning for a young morphologist then-as well as now-because it called into question what Gould and Lewontin (1979) were later to call the "adaptationist paradigm."

A few years later Alberch et al. (1979) reformulated Gould's clock model into a set of Cartesian diagrams. These were far simpler than the original clock-model diagrams and succeeded in joining the topics of allometry and heterochrony within a unified and mathematically elegant framework. This article also provided Gould with an opportunity to relate heterochrony to his emerging theory of macroevolution as a possible mechanism for saltatory phenotypic shifts and trans-specific trends in phenotypic evolution (the later coming about via the operation of species selection). Then Bill Fink (1982) stressed the need to formulate tests of heterochrony within the context of formalized phylogenetic hypotheses, thus placing this research program firmly within the emerging juggernaut of phylogenetic systematics. In particular, Fink (1982) showed that 1) not all morphological transition patterns need be thought of as heterochronic (e.g., some could be retained plesiomorphies), and 2) under some scenarios it might not be possible to determine whether heterochronic or non-heterochronic patterns were present in comparative datasets without access to additional sources of information.

Throughout the 1980s and early 1990s the study of heterochrony took off as a major (paleo)biological research program. This program culminated with publication of M. L. McKinney and K. J. McNamara's books (McKinney, 1988; McKinney and McNamara, 1991), and especially with McNamara's (1997) book where it is claimed that heterochrony permeates every "nook and cranny of evolution. . . without it evolution wouldn't have happened. . . it explains everything from the shape of a delphinum flower, to a horse's foot, to the song of a bird" (p. 46). As was perhaps inevitable, the pendulum had swung too far. Heterochrony had become overextended as an explanation for morphological phenomena. A correction was needed to redress the imbalance.

The arguments against heterochrony's hegemony came in two types and they came in the same year. The strong argument was advanced by Rudolf Raff in his 1996 book The Shape of Life: Genes, Development, and the Evolution of Animal Form. Therein Raff noted that heterochrony is grounded on the assumption that morphological change arises because quasi-autonomous units of an organisms' developmental program can become dissociated from one another and, once dissociated, modifiable via alterations in only three factors: the offset signal, the onset signal, and the rate of development, all of which are ways of altering the relative timing of somatic reproductive development. These three types of variation, coupled with the two alternative modes of heterochronic change-peramorphosis: somatic development speeded up relative to reproductive development; and paedomorphosis: somatic development slowed relative to reproductive development-yielded the six basic types of heterochrony (Fig. 1). Since any ontogenetic trajectory can be represented in the Cartesian space defined by these (conceptual) axes of reproductive and somatic development, and since any linearized set of ontogenetic trajectories can be described by appropriate modifications of the onset, offset, and rate parameters, this representational system can be used to describe any conceivable morphological transition. However, does this mean that that the inference of all biological process effecting morphological change can proceed in a unambiguous manner from the simple documentation of reproductive/somatic pattern? Raff thought not and drew on his own work as well as that of others to explain how to test the implicit pattern-process link necessary to identify heterochronic processes correctly. Raff's review with G. A. Wray (1989) on Axolotl development-often cited as a classic example of paedomorphosis (see Gould, 1977)-makes this point nicely.

The weaker criticism of heterochrony was represented by Zelditch and Fink's (1996) work on heterotopy as an additional factor in the description of development. These authors didn't see heterochrony as an incorrect hypothesis of evolutionary development, so much as an incomplete hypothesis; incomplete because it focused attention on the temporal domain of an irreducibly spatiotemporal plexus. To remedy this incompleteness Zelditch and Fink sought to reinstate Heackel's original concept of heterotopy (change in the position of a structure or growth field over the body) as a necessary partner to heterochrony in ontogenetic pattern descriptions. I regard this as the weaker criticism because, unlike Raff's more process-oriented work, it only concerns itself with the comparison of patterns. Thus, heterochronic and heterotopic patterns can be documented, but this documentation, by itself, givs no indication whether the developmental processes responsible for those patterns are themselves heterochonic or heterotopic. Zelditch and Fink's (1996) strict focus on pattern analysis also carried with it the seeds of its own destruction in much the same way as McNammara's pan-heterochrony stance. The point of Gould's (1977) book that so invigorated this field was not to formulate a novel or more holistic way of describing morphological change. Rather it described a new class of putatively process-related evolutionary phenomena that could, in principal, better explain the origins of such changes. Zelditch and Fink's (1996) take on heterotopy appeared to come out of a purely descriptive tradition that owned more to conceptual developments in applied geometry than to conceptual advances in evolutionary theory.

Coming more than half a decade after these works I anticipated that Zelditch's book Beyond Heterochrony: The Evolution of Development would summarize what's been going on in this field in the interim. I was especially looking forward to new information about the way that the new 'evo-devo' research program was making progress in the exploration of developmental systematics via comparative analysis within a formal phylogenetic-morphologic context. I was hoping that some concept transfers from other morphology-based research programs would have been made by now (e.g., "comparative method" analysis, see Harvey and Pagel, 1991; Harvey et al., 1996), and I was anticipating at least a few attempts at a synthesis between developmental genetics and morphological analysis that focussed on the evolutionary dynamics of such systems.

The book begins with a rather generic Foreword by Brian Hall. On first inspection, this piece seems to do little more than note that people currently disagree with the role of heterochrony in evolution and introduce the concepts of heterotopy, heterotypy (change of type), and heterometry (change of amount) as the alternatives to heterochrony implied by the book's title. Hall claims that these, as a group, are the real sources of novelty in evolution.