The ability to survive in a dormant state is a widespread, yet unevenly distributed feature among invertebrates. Organisms belonging to classes and phyla that include fresh-water or terrestrial representatives are more likely to possess a dormant stage than those in groups that are exclusively marine. Moreover, within taxa where has evolved, it is more common among fresh-water and terrestrial species than in marine species. This correlation between and habitat across 29 free-living invertebrate phyla raises the question of cause and effect. Are dormant stages more commnon in fresh-water and terrestrial habitats because there is greater selection for in those environments, or is a prerequisite for the successful invasion of non-marine systems? The mechanism of varies among species and ranges from a specialized diapausing embryo to a quiescent adult. Dormancy has most likely arisen multiple times in invertebrate life histories, both within and between phyla. Although may facilitate the invasion of fresh-water and terrestrial habitats, it is not always a requirement: many non-marine species do not exhibit dormancy. Additional key words: diapause, quiescence, life history, fresh-water, marine All natural environments vary, and organisms may occasionally experience conditions that are limiting for growth, survival, or reproduction. The form and nature of this variability differ among habitats, however, and this has led to the evolution of a wide array of life histories (reviewed by Roff 1992; Stearns 1992). For example, some species are iteroparous and offset the negative effects of temporal variation with long-lived adults that reproduce multiple times (Murphy 1968; Goodman 1984). Alternatively, other organisms avoid deleterious environmental conditions with reliable migration (Dingle 1978; Levin et al. 1984; Wiener & Tuljapurkar 1994). However, many species without either of these traits persist in variable environments. Often, these taxa inhabit relatively isolated, transient systems that can support life only for a fraction of the year (e.g., arctic tundra, deserts, temporary ponds). When organisms cannot migrate away from declining environmental conditions, and active adults cannot survive, the only viable option is persistence through dormancy. Dormancy has evolved in numerous bacterial, fungal, protist, plant, and animal species, but is completely unknown in many groups. Although all forms result a Present address: Illinois Natural History Survey, Champaign, IL 61820, USA. Voice: 217-244-2139. Fax: 217333-6294. E-mail: caceres@mail.inhs.uiuc.edu in some type of metabolic and/or developmental depression, the term dormancy actually encompasses a wide range of physiological states (see Hand 1991). Therefore, the form and duration of the dormant state vary widely among taxa. Whereas some species remain dormant only as long as conditions are unfavorable, others remain dormant for time periods much longer than the unfavorable environmental conditions, in :some cases for a century or more (Harper & White 1974; Henis 1987; Hairston et al. 1995; Caiceres 1997). ]Dormancy that extends beyond the duration of the environmental hardship introduces potential costs and benefits. When is not terminated upon the return of favorable conditions, there is the potential cost of lost reproduction. However, when the dormant prcpagules from a single cohort continue to introduce recruits into the active population for several years, populations of short-lived adults can sample different environments much in the same way as a long-lived iteroparous species (Seger & Brockmann 1987). Thus, overlapping generations are established in populations that might otherwise consist of discrete cohorts (Templeton & Levin 1979). ]In theory, the evolution of prolonged should be an alternative to either iteroparity or dispersal (Venable & Lawlor 1980; Levin et al. 1984; Rees 1994). In fact, these alternative life histories have been found in some natural systems. Rees (1993), in This content downloaded from 157.55.39.127 on Wed, 29 Jun 2016 04:32:53 UTC All use subject to http://about.jstor.org/terms