Publication Date



Open access

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Biology (Arts and Sciences)

Date of Defense


First Committee Member

Carol C. Horvitz

Second Committee Member

Donald L. DeAngelis

Third Committee Member

J. Albert C. Uy

Fourth Committee Member

R. Stephen Cantrell

Fifth Committee Member

Bette A. Loiselle

Sixth Committee Member

Paul D. Pratt


Multiple introductions of an exotic species can facilitate invasion success by allowing for a wider range of expressed trait values in the adventive range. Schinus terebinthifolius Raddi (Brazilian Pepper) is an invasive shrub that was introduced into Florida in two separate introductions and has subsequently hybridized, resulting in three distinct lineages (eastern, western and hybrid). In the first chapter, to determine whether allocation of above ground biomass differed by introduction history we destructively sampled 257 stems from each of 6 populations with differing introduction histories. The proportion of aboveground biomass allocated to fruit, wood and leaves differed among the three populations. To determine whether the relationship between stem size and several dependent variables that measure plant performance (total dry weight, wood dry weight, number of fruits, fruit dry weight, leaf dry weight and number of leaves) differed quantitatively by introduction history, we performed analyses of covariance. Slopes of these relationships (dependent variable vs stem size) varied by lineage. Hybrid populations had the steepest slopes for one set of dependent variables (total dry weight, wood dry weight, and leaf dry weight), while Western populations had the steepest slopes for a different set of dependent variables (number of fruits, fruit dry weight and number of leaves). Schinus has bright red berries that are dispersed by animals during the winter months when Audubon Society Christmas Bird counts take place. In the second chapter we used hierarchical cluster analysis with sixteen years of bird abundance data at six sites to assess spatiotemporal patterns in the assemblage of potential dispersers. We found that the assemblage of potential avian consumers of Schinus was more variable between sites than between years. Long Pine Key had the most distinct assemblage. Ft. Myers had the highest percentage of potential seed predators, followed by Ft. Pierce and Cocoa. While we examined several environmental variables including temperature, precipitation, landcover and daylength, none showed patterns of variability consistent with the spatiotemporal patterns of variability in the avian frugivore communities at the six sites in the sixteen years. Thus, while sites varied in the composition of their avian frugivore communities across space and time, it remains unknown which, if any, of the environmental factors is causative of the spatiotemporal patterns we observed. In the third chapter, we observed avian visits to patches of reproductive Schinus at three locations across Florida to identify the community of avian dispersers in Florida and to estimate removal rates. We observed adult reproductive female Schinus trees at three sites (Punta Gorda, Palmetto Bay Village Center and Ft. Pierce) and recorded all sightings and removal events during the four-hour period just after dawn for 47.5 person hours. We observed American Robins, Grey Catbirds, Northern Mockingbirds and Pine Warblers removing Schinus seeds. Overall, most of the removal events we observed were by Grey Catbirds. The rate of removal events per infructescence was highest at Ft. Pierce, followed by Palmetto Bay Village Center, while the rate of removal events per infructescence at Punta Gorda was the lowest. Finally, in the last chapter we construct an integral projection model for Schinus. We modeled survival, growth and reproduction as functions of two continuous variables– the diameter at the base of the plant and the height of the plant–on two continuous size domains. We found that small individuals had the highest mortality, while very few large individuals experienced mortality. We first constructed an integral projection model using the pooled data across all sites. The population growth rates we calculated were all above one, indicating that the population is increasing. The elasticity analysis revealed that the largest elasticity is in the survival and growth of the largest individuals, suggesting that management actions that target the largest individuals should have the greatest impact on the population growth rate. We also constructed three additional integral projection models–one for each biotype (Eastern, Hybrid and Western). The Western biotype had the highest population growth rate, and the results of a LTRE analysis separating the contributions of each vital rate to the differences in population growth rate revealed that the Western biotype’s increased population growth rate came mainly from higher probabilities of graduating from the smaller size domain into the large size domain than either the Eastern or Hybrid biotypes.


Schinus terebinthifolius; invasive species; integral projection models; dispersal; biomass allocation; frugivores