Wertz, Parker (Weber State University)
Faculty Advisor: Dorsey, Bryan (Weber State University, Geography)
Prevention and predicting spread is the best method of control against invasive species. Land managers require accurate and reliable methods for containment and eradication to prevent land cover change and loss of biodiversity. Ecological niche models exist and are used by ecologists to map habitat suitability, but many rely on presence-absence samples which are difficult to obtain. Maximum entropy species distribution modeling (Maxent) is a popular model that has been increasingly used since it can make valid predictions using presence-only data. Many studies have used Maxent to model species distributions, but few have done so with crowdsourced data since it is more likely to be bias and unreliable. The purpose of this study is to test the robustness of Maxent using crowdsourced presence-only data on Lyrthum salicaria, a perennial herb that invades wetlands and pushes out native flora. The study is set in northern and central Utah, and uses environmental variables in climate, landcover, and topography, with landcover being the most contributive factor to the model. Model performance was very good, even with species data being bias towards areas of higher population, proving Maxent as a worthy method to use in species distribution modeling with crowdsourced species presence data. This results of this study show promise for use in modeling other invasive plants in the future.