JOURNAL OF EXPERIMENTAL ZOOLOGY (MOL DEV EVOL) 288:1–20 (2000) Plasticity and Constraints in Development and Evolution JASON HODIN* Science and Math, Seattle Central Community College, Seattle, Washington 98122 ABSTRACT Morphological similarities between organisms may be due to either homology or homoplasy. Homologous structures arise by common descent from an ancestral form, whereas homoplasious structures are independently derived in the respective lineages. The finding that simi- lar ontogenetic mechanisms underlie the production of the similar structures in both lineages is not sufficient evidence of homology, as such similarities may also be due to parallel evolution. Parallel- isms are a class of homoplasy in which the two lineages have come up with the same solution independently using the same ontogenetic mechanism. The other main class of homoplasy, conver- gence, is superficial similarity in morphological structures in which the underlying ontogenetic mecha- nisms are distinct. I argue that instances of convergence and parallelism are more common than is generally realized. Convergence suggests flexibility in underlying ontogenetic mechanisms and may be indicative of developmental processes subject to phenotypic plasticity. Parallelisms, on the other hand, may characterize developmental processes subject to constraints. Distinguishing between ho- mology, parallelisms and convergence may clarify broader taxonomic patterns in morphological evo- lution. J. Exp. Zool. (Mol. Dev. Evol.) 288:1–20, 2000. © 2000 Wiley-Liss, Inc. As the fields of developmental and evolution- lar approach. I argue, on the contrary, that since ary biology continue to converge, an underlying innovations are manifest at the morphological pattern is beginning to emerge: namely that the level, it is necessary to integrate a morphological astonishing diversity of morphological variation with a molecular approach if we hope to uncover in plants and animals is built on a scaffolding of any such underlying principles. Specifically, I feel a seemingly quite limited set of developmental that the study of incidents of parallel and conver- programs. This pattern was predicted at least as gent evolution may offer the best approach to ac- far back as the mid-1970s by Emile Zuckerkandl complish this integration. (’76), at which time he wrote: Before continuing, I now provide definitions for several key terms. “Parallelism” is independent Functional innovation at the morphologi- evolution using the same mechanism. This is con- cal level does not appear to require any func- trasted with “convergence”: independent evolution tional innovation at the molecular level. One using an alternate mechanism. The third term in may wonder whether the formation, in the this class is “reversal,” or a return to an ances- course of evolution, of a limb of a land verte- tral mechanism (a special case of parallelism). brate from the fin of a fish requires the ap- “Homoplasy” is a broader term covering parallel- pearance of new proteins endowed with novel ism, convergence and reversal: similarity not re- functions. I am inclined to believe that it does sulting from common ancestry. The definition of not. (p 404) “homology” is certainly more contentious (re- viewed in Donoghue, ’92; Abouheif et al., ’97). I So when we catch our collective breath from tentatively adopt the definition of Van Valen (’82): “startling” findings such as the apparent conser- vation of Hox and Pax gene functions, let us step back for a moment and consider the implications. There are optimists who feel that since the set of Grant sponsor: National Institutes of Health; Grant number: developmental mechanisms appears to be limited, HD07183. we are likely to come to an understanding of the *Correspondence to: J. Hodin, Science & Math, Seattle Central Community College, 1701 Broadway, Seattle, WA 98122. basic nature of metazoan development and evolu- E-mail: hodin@ alumni.washington.edu tion via a reductionist developmental and molecu- Received 14 August 1999; Accepted 14 September 1999 © 2000 WILEY-LISS, INC. 2 J. HODIN “resemblance caused by a continuity of informa- various tiger salamanders (Shaffer and Voss, ’96), tion.” This definition highlights an important dis- when salamanders as a whole are considered, dif- tinction between homology and homoplasy, as the ferent steps along the thyroid hormone response latter involves a definite historical discontinuity axis are blocked in different paedomorphic spe- of information. We should proceed with the real- cies (Frieden, ’81; Yaoita and Brown, ’90), leading ization that in practice the distinction between to a similar metamorphic failure. Thus although homology and homoplasy is often far from clear the genetic bases of the paedomorphic phenotype and that other definitions of homology (such as are clearly different (i.e., convergent at the genetic the “building block hypothesis” of Wagner, ’95) level) in different salamander species, similar de- may often prove more useful as heuristic devices. velopmental mechanisms (at the level of the gene The term “mechanism” in the context of con- network) appear to be involved, so I consider these vergence and parallelism is in need of some clari- disparate cases of paedomorphosis to have evolved fication. In an entirely reductionist approach, “the in parallel. In order to learn how development may same mechanism” refers to identical genetic bases. influence evolution, we should focus on similari- For example, paedomorphosis (reproduction at an ties and differences at the level of the develop- immature morphological stage) has evolved sev- mental mechanism (rather than, say, the genetic eral times independently in tiger salamanders (re- level1) when making evolutionary comparisons. viewed in Shaffer and Voss, ’96). Under the strict In Part I of this paper, I concentrate on some reductionist view, these independent instances of high profile examples in the literature where ho- paedomorphosis are considered to have evolved in mology has been cited as the cause of similarity parallel only if they are characterized by muta- between arthropods and chordates. It seems that tions in the same gene or genes. Yet, discovery of the possibility that these similarities are due to the genetic bases for homoplastic changes in many parallel or convergent evolution has been dis- organisms is not practically feasible. Thus, partly missed too readily. I argue that the apparent for practical reasons, I adopt a developmental defi- underestimation of the role of homoplasy in met- nition of mechanism. If the cell biological and mor- azoan evolution skews our perspective on how de- phogenetic changes are identical, then I consider velopment evolves. In Part II, I explore the it to be an example of a parallelism. Furthermore, possibility that the study of well-documented cases since developmental mechanisms often appear to of homoplasy not only yields useful information be characterized by dense networks of cross-regu- on evolutionary patterns, but also offers signifi- latory and feedback interactions among genes, cant insights into the ways in which interacting changes in several different members of a given networks of regulatory genes work to produce com- gene network could produce identical morphoge- plex morphological structures. I distinguish be- netic phenotypes. I consider such examples to be tween cases of parallel evolution, which may parallel evolution at the level of the developmen- indicate the presence of developmental constraints tal mechanism. In the case of the tiger sala- (a bias in the production of phenotypic variation manders, precocious sexual development versus due to ontogenetic factors; Maynard-Smith et al., delayed adult development represent different de- ’85), from instances of convergent evolution, which, velopmental mechanisms leading to a similar pae- I hypothesize, correlate with phenotypic plasticity domorphic phenotype and, therefore, an example (environmentally induced alterations in morphol- of convergent evolution at the level of the devel- ogy within a genotype) in underlying developmen- opmental mechanism. Alternatively, all tiger sala- tal mechanisms. manders might be characterized by delayed adult development via blocks in the metamorphic path- PART I: HAS NOTHING INTERESTING way. Even if these blocks were at different points HAPPENED SINCE THE CAMBRIAN? in the metamorphic pathway (which would be an With the advent of DNA sequence analysis, the example of convergence at the genetic level, since word “homology” is suddenly in every molecular mutations in different genes would underly the biologist’s vocabulary. References to “the verte- instances of homoplasy), the developmental tra- brate homolog” of, say, a Drosophila gene are not jectories would be similarly altered and thus only commonplace but, indeed, are becoming the would represent a case of parallel evolution at the level of the developmental mechanism. Although 1Still, considering homplasy at the genetic level is useful in other the specific developmental mechanisms leading to contexts, such as understanding the evolution of protein function (see paedomorphosis are poorly understood among below). PLASTICITY AND CONSTRAINTS 3 cornerstone of discourse in modern developmen- nity. What followed was a flurry of conclusions tal genetics. In addition to the inherent problems (not hypotheses or suggested experimental tests) in using the same term to describe similarity at that the deuterostome-protostome ancestor had two different levels (morphological