Predicting the spatial distribution of an invasive plant species and modeling tolerance to herbivory using Lythrum salicaria L. as a model system

Understanding the ecological factors behind the landscape-level distribution of invasive species is a rapidly growing area of research with strong applied implications. In a major part of my thesis, which comprises chapters 2 &3, the focus is on spatial pattern analyses and predictive modeling of an invasive wetland plant: Purple Loosestrife (Lythrum salicaria L.). More specifically, the first part of my thesis (i.e. chapter 2) considers a novel hierarchical approach, wherein the spatial distribution of loosestrife in a human-modified landscape was found to be the consequence of three key hierarchical factors: wetland habitat availability, disturbance prone surrounding land-use conditions around the wetland habitat, and propagule pressure. In chapter 3, the spatial factors and ecological processes characterized in chapter 2 were put-together and several logistic and autologistic regression models were developed to predict locations of loosestrife occurrences. Incorporating propagule pressure as an autocovariate was found to be crucial in making accurate predictions of loosestrife invasion risk. However, in the absence of propagule pressure, the surrounding land-use model highlighted the role of anthropogenic edges in defining the invasibility of wetland habitats. From an applied perspective, the model based risk maps assist conservationists and land managers in predicting and checking the spatial spread of invasive loosestrife. In the fourth and last research chapter of my thesis, a mathematical model is developed to explore herbivore tolerance in perennials with long-term belowground storage. The inspiration behind this model is loosestrife, an invasive perennial, and its biocontrol insect herbivores. More specifically, a discrete time model was built to explore the role of belowground allocation of biomass in a perennial plant with distinct growing season and under regular seasonal defoliation by herbivores. The model addresses the role of two co-occurring traits like utilization of stored reserves for early-season growth and post-herbivory regrowth and consequent tolerance potential. The model results highlighted that belowground biomass allocation plays a critical role as it allows the plant to persist despite severe periodic defoliation by herbivores. The model findings also indicated that when highly efficient early-season use of stored reserves is coupled with high belowground biomass allocation potential the plant biomass and herbivore population can show sustained cycles. From the perspective of invasive perennials, the model suggests that brief periods of intense seasonal herbivory is incapable of extirpating the invasive plant population as long as the latter can efficiently allocate biomass belowground