898 Ecology, 83(4), 2002, pp. 898–905  2002 by the Ecological Society of America EFFECT OF COMMUNITY STRUCTURE ON INVASION SUCCESS AND RATE THOMAS E. MILLER, 1 JAMIE M. KNEITEL, AND JEAN H. BURNS Department of Biological Science, Florida State University, Tallahassee, Florida 32306-1100 USA Abstract. Although invasion has long been recognized as an important ecological pro- cess, there are very few experimental studies of invasion in natural communities and vir- tually no studies that determine how trophic structure affects the probability of invasion. We introduced novel protozoans and rotifers into the natural communities found in the water-filled leaves of the pitcher plant Sarracenia purpurea. The communities were ma- nipulated in a factorial design of removal of predators (larvae of the mosquito Wyeomyia smithii) and addition of resources (dead insects). Three of the six protozoan species suc- cessfully established populations when introduced into pitchers, suggesting that these spe- cies are migration limited. The other three protozoans and a rotifer did not successfully invade established communities, although all four are naturally found in these inquiline communities. Of the three successfully invading protozoans, two were more likely to invade when resources were added and one of those even more frequently when predators were removed. Invasion by the third was unaffected by these experimental manipulations. Similar effects of predators and resources were found on population sizes of these three species. This study is one of very few that have addressed invasion experimentally; its results suggest that a variety of factors, including migration, predation, and resource availability, can have different influences on invasion by fairly similar protozoans. Key words: aquatic community invasion, manipulative study of; community structure affects invasion success; Habrotrocha rosa; invasion success, factors affecting; predation; Protozoa; resource levels; Sarracenia purpurea; Wyeomyia smithii. INTRODUCTION Invasion by alien species has increasingly been rec- ognized as one of the major worldwide environmental problems of the last century. ‘‘Invasion,’’ or migration of new species into local communities, can also be important for community development (e.g., succes- sion, Connell and Slatyer 1977) and for community structure and composition (MacArthur and Wilson 1967, Ricklefs and Schluter 1993, Loreau and Mouquet 1999). Ecologists know remarkably little, however, about what factors facilitate or prevent invasion, in part because of the dearth and difficulty of experimental studies. Most of our current knowledge of this process comes from either observational or theoretical studies. Ob- servational studies have focused on identifying traits that are common to successful invasive species (New- some and Noble 1986, Noble 1989, Perrins et al. 1992, Veltman et al. 1996). It has frequently been assumed, for example, that successful invasive species will have traits of ‘‘weedy’’ species (Baker 1974), such as rapid growth to reproductive age, high offspring production, and potential long-distance dispersal of propagules. It has proven difficult, however, to categorize successful invasive species by any unique set of traits (see, e.g., Newsome and Noble 1986, Lonsdale 1994). Manuscript received 14 September 2000; revised 10 July 2001; accepted 13 July 2001. 1 E-mail: miller@bio.fsu.edu Theoretical approaches have emphasized traits of communities that affect the probability of invasion, rather than traits of invasive species. This work has largely developed from Elton’s suggestion that resis- tance to invasion is correlated with species number (Elton 1958, Levine and D’Antonio 1999), an idea that is certainly consistent with, if not drawn from, the niche concept. Several theoretical models have proposed a negative relationship between the probability of suc- cessful invasion and species richness or between the probability of successful invasion and the intensity of species interactions (Case 1990, 1991, Cornell and Lawton 1992, Morton and Law 1997). In these models, the dominant force limiting invasion is competition for resource niches, and current occupants are generally viewed as having a competitive advantage. Successful invaders must generally displace current species or use empty or underused niches. Although several studies have investigated invasion using laboratory microcosms (e.g., Robinson and Dick- erson 1984), very few studies have used controlled experimental introductions of species into natural com- munities (see Levins and Heatwole 1973). Several stud- ies have investigated the effects of native species rich- ness on invasion of natural communities, with contra- dictory results. Whereas some studies found that spe- cies richness decreased the probability of successful invasion (e.g., Tilman 1997, Stachowicz et al. 1999), others found no effect of richness (e.g., Peart and Foin