Predicting The Effects Of Policies To Control The Spread Of Forest Insects And Diseases And To Eliminate Illegal Logging PDF Download

"The primary goal of this thesis was to identify largescale generalities across species invasions. Arguably, due to the dominant role of humans, many of the drivers of invasive species establishment, spread, and associated damages should be common among species. However, whether generalized predictive models were possible or whether species must be modelled and fit separately to account for idiosyncrasies remained unanswered. In Chapter 1, I built a single, general model for the spread of all invasive forest pests in the United States that explained more than 75% of the variation in pest distributions. In Chapter 2, I contrasted this general model with more complex single-species models. I also generated semi-generalized models using the general model as a starting point and adding simple species-specific correction factors. I found that single-species models performed 17% better than the general model but underperformed compared to semi-generalized models that incorporated cross-species generalities. In Chapter 3, I tackled pest establishment and its relation to host diversity. Many current hypotheses (e.g., dilution and amplification effects) have been considered in isolation, have heterogeneous support, and can be strongly affected by a range of confounding factors. I developed a general model, which simultaneously integrated processes argued in the literature, to allow a more rigorous, systematic analysis of biodiversity effects on pest occurrence. I measured host diversity as phylogenetic dissimilarity from the focal host community of a particular pest, and modelled it simultaneously with pest spread patterns, host richness, and host density effects. I detected a nonlinear relationship, where positive density dependence emerged for closely related species (amplification), and negative density dependence (dilution) emerged for more distantly related species. In Chapter 4, I projected future damages to urban trees by synthesizing my pest invasion forecasts with models of urban tree distributions and tree mortality. I projected future urban tree mortality of roughly 385M trees over 30 years (including 8% of all street trees). I provided updated estimates of the economic costs of this mortality, advancing the field by 1) leveraging newly created, more complete datasets, 2) providing increased spatiotemporal resolution, and 3) separating the costs across pest species causes, host tree genera, and urban tree categories: street, residential and community. Taken in sum, these models produce general predictions across invasion stages that can act as inputs to invasive species control programs, and I plan to use them in subsequent work to develop general heuristics for effective pest management"--