Several years ago, I suspended laboratory research on development in Cnidaria in favor of library research on the evolution of development. I dug deeply into the literature (back to Aristotle), discovered some little known work and rediscovered some well-known texts. My belief that development is rooted in evolution was reinforced on every page, and, in an effort to share what I had learned, I wrote Embryology: An introduction to developmental biology (New York, Harper-Collins, 1991), a textbook on animal embryology written from a comparative point of view.
My research did not stop there. A variety of questions had occurred to me during the writing which I picked up afterward, again, in the library. The first question was, "Are the mechanisms and controls operating during asexual reproduction different from those operating during sexual reproduction?" I chose to answer this question for the Cnidaria, since I was already acquainted with several members of this phylum. Cnidarians are also suitable subject matter, since many exhibit both forms of reproduction. The answer, presented in detail in "Studies of asexual reproduction in Cnidaria," is that the mechanisms and controls of sexual and asexual reproduction only begin to correspond at terminal differentiation.
In retrospect, this answer is not surprising, since asexual reproduction is a phenomenon of epithelia, while sexual reproduction is a phenomenon of amoeboid cells. Epithelia is a steady-state type tissue, while amoeboid cells exhibit stem cell dynamics. Other differences between epithelia and amoeboid cells are documented in the experimental literature of both Hydra and larval marine hydroids. Epithelial cells alone give rise to epithelia and muscle, while amoeboid cells (also known as interstitial cells) alone give rise to cnidocytes (the sources of cnidocysts), nerve, sensory cells as well as egg and sperm. In a report entitled "A symbiogenetic theory for the origins of cnidocysts in Cnidaria," I argue that the separation of function which accompanies the separation of cnidarian epithelia and amoeboid cells coupled to the reintegration of function following recombination resembles the relationship of a host to a parasite. I suggest, therefore, that epithelia and amoeboid cells in Cnidaria have separate origins: the epithelia being derived from an exclusively epithelial ancestor, possibly resembling extant Tricoplax; the amoeboid cells being derived from one or more parasitic protoctistan, possibly resembling extant members of the Microsporidia.
In order to pursue this hypothesis, I decided to trace the evolution of cnidocysts, since a large literature, going back nearly a century, already exists on their distribution in different taxa. With the assistance of interlibrary loan, I accumulated comprehensive data on the qualitative and quantitative distribution of thirty or more types of morphologically defined cnidocysts in nearly a thousand species. These data are now available to the public as a cnidocysts database. My paper, "Origins of diversity in cnidarian cnidocysts," which introduces the cnidocyst database, describes a startling finding: morphologically identified types of cnidocysts are traced back to at least two different sources in Cnidaria's subphylum known as Medusozoa, while cnidocysts are traced back to only one source in Cnidaria's other subphylum known as Anthozoa. Differences in the range of cnidocyst diversity, moreover, suggest that competitive exclusion and character displacement influenced cnidocyst evolution in Medusozoa. The possibility arises, therefore, that different evolutionary mechanisms came to bear on the evolution of differentiation in various tissues in multicellular organisms. In the future, I intend to pursue this possibility by characterizing the quantitative data in the cnidocyst database, and thus exploring quantitative evolution, and by refining my guess as to the originary sources of cnidocysts by looking further into the literature on candidate protoctistans.