Int. J. D BioI. ~O: 859-869 (1996) 859 Direct development in the lungless salamanders: what are the consequences for developmental biology, evolution and phylogenesis? DAVID B. WAKE' and JAMES HANKEN" 'Museum of Vertebrate Zoology, and Department of Integrative Biology, University of California, Berkeley and 2Department of Environmental, Population, and Organismic Biology, and University Museum, University of Colorado, Boulder, USA ABSTRACT Direct development is a widespread alternate reproductive mode in living amphibians that is characterized by evolutionary loss afthe free-living, aquatic larval stage. Courtship, mating, and oviposition occur on land, and the terrestrial egg hatches as a fully formed, miniature adult. While it is the most common reproductive mode in urodeles, development outside the reproductive tract of the female that proceeds directly to a terrestrial hatchling occurs in only a single lineage, the lungless salamanders of the family Plethodontidae. Evolution of direct development in plethodontids has contributed importantly to the extraordinary evolutionary success of this speciose, geographically widespread, and morphologically and ecologically diverse taxon. Developmental consequences and correlates include increased egg size and embryonic development time, loss of larval structures and ontogenetic repatterning, and altered pattern formation in organogenesis. Evolutionary and phylogenetic consequences and correlates include the loss of larval constraints and origin of morphological novelty, and frequent homoplasy. Analysis of direct development in an evolutionary context illustrates the complex interplay between processes of phylogenetic divergence and develop- mental biology, and substantiates the prominent role of developmental processes in both constrain- ing phenotypic variation and promoting phenotypic diversity. Despite the proven suitability of direct- developing plethodontid salamanders for laboratory and field study, knowledge of basic features of their developmental biology remains far below that available for many other urodeles. Examination of such features of these "non-model" organisms is an appropriate and deserving goal of future research. KEY WORDS: direct develojJment, PletllOdontidae, evolution, onlogen)', constraints Introduction Salamanders have played an indispensable role in the history of developmental biology. The vast majority of studies of urodele development involve species in which embryogenesis culminates in the production of a free-living larva. In most of these taxa, the larva, after a period of weeks or months, metamorphoses to a terrestrial adul!. Yet, this complex life history, which most people regard as a characteristic if not defining feature of living amphib- ians, is not the predominant reproductive mode in urodeles. Instead, most living salamanders have direct development: court- ship, mating, and oviposition occur on land. and the terrestrial egg hatches as a fully formed, albeit miniature adult; there is no free. living larva (Figs. 1,2). Although it is the most common reproductive mode in salaman- ders and has evolved independently in frogs and caecilians (Wake, M., 1989), development outside the reproductive tract of the female that proceeds directly to a terrestrial hatchling without a free-living larval stage is restricted to a single urodele lineage, the lungless salamanders of the family Plethodontidae. Direct development in plethodontids is believed to underlie, at least in part, the extraordi- nary evolutionary success of this family, which is manifest both in the large number of living species and in their extensive morpho- logical, functional, and ecological diversity (Wake and Larson, 1987; Wake. 1991). In this paper, we summarize the biology a[ direct development in plethodontids, focusing on the numerous consequences and correlates of this derived reproductive mode both fordevelopmental biology and forevolution and phylogenesis. Most studies of direct development in plethodontids have focused on ecological and evolutionary aspects: there have been few analyses of development per se. As a means of encouraging more studies of the developmental biology of direct development in iO Address for reprints: Departmentof EPQ Biology, University of Colorado, Boulder, CO 80309-0334, USA. FAX: 303-492-8699. e-mail: James.Hanken@Cotorado.edu 0214-6282196/$03.00 e liRC PT(:S~ Printed in Sp~;n -- -- --