| Large, abrupt changes in communities and ecosystems are widespread in nature and have major consequences for ecosystem services and human wellbeing. Sudden changes can occur if a system responds to a driving, environmental variable linearly, nonlinearly with a tipping point, or in a complex, nonlinear manner, with multiple equilibria over a range of the environmental driver (i.e., alternative states). The goal of my dissertation was to identify the characteristics of systems that make them prone to a tipping point. I used multiple lines of evidence (field surveys, lab experiments, and simulation modelling) to understand how multiple levels of biological organization -- from individual species and their traits to the physical properties of ecosystems -- interact to determine the presence of and shift between ecological states. I used freshwater lakes and ponds, which are often dominated by either floating- or submerged-plants, as a model system. In the northeast United States, I found that the floating plant state was limited by the size of the water body (< 5 ha) and the absence of floating plants from approximately 60% of lakes and ponds. In small ponds where floating plants occurred, there was evidence that floating plant dominance is an alternative state, driven by nutrient levels in the water. In laboratory experiments, species of floating plants exhibited unique growth and dormancy performance in response to environmental conditions (temperature and nutrient stoichiometry). Although polycultures of floating plants were expected to be more dominant, there was only weak evidence for this pattern in the field. To integrate field and lab results, I developed a spatially- and temporally-explicit simulation model. I found that the model produced alternative states, where simulations ended in either a floating- or a submerged-plant state, depending only on the initial conditions. In model simulations, floating plants were less dominant in larger water bodies, as expected from field surveys, if wind strength increased with water body size and had a prevailing direction. Floating plant species diversity or trait composition did not increase the dominance of this group. This work shows that the tipping point between ecological states is dependent on multiple factors. Includes supplement data file ("sample_input.csv") for running the simulation model. |
