Multispecies interactions and demography of the invasive thistle Centaurea solstitialis in California

Plants interact with numerous other species, some (e.g., pollinators) increase plant fitness while others (e.g., herbivores) decrease it. Because individual performance can be decoupled from population performance, even interactions that have large effects on individual plants may not affect the plant's local abundance or population growth rate. In general, we know very little about multispecies interactions and under what circumstances they affect the plant populations. My dissertation examines how a suite of interacting species affects both individual Centaurea solstitialis plants and their populations in California. C. solstitialis is native to Eurasia and was introduced to coastal California over a century ago. It has since spread to nearly every habitat type in the state. I explored how interacting with multiple species affects (1) individual plant performance and (2) the pathways by which those affects arise, and (3) under what circumstances these interactions have population-level effects. I factorially manipulated the presence of interacting species in field settings arranged latitudinally across the invaded range and found that plants attacked by multiple enemies did not produce fewer seeds than plants attacked only by one enemy. But I also found that the lone metric of seed production masked both superadditive and subadditive indirect interactions. I used both projection matrix models to describe the dynamics of established invasions at all of my study sites and a seed addition experiment to assess the relationship between seed production and plant density and the long term trajectory of each population. I found that in some populations reducing seed input reduced the density of flowering plants but this relationship was much weaker in other populations. I found strong effect of flowering plant density on seed production such that even when seed input lead to lower flowering plant density it did not reduce total seed produced, so the long-term trajectory of the population remained stable. Taken together my results indicate that interactions among species can be complex and produce unexpected outcomes for the shared host plant C. solstitialis and that reducing seed input may be an effective approach to managing some but not all C. solstitialis invasions in California.