Big data for the benefit of aquatic invasive species management

Aquatic invasive species (AIS) are an expensive problem and the best thing we can do to manage AIS is to prevent their spread. Preventing the spread of AIS requires an interjurisdictional and big data approach to find pathways, locations, and people to prioritize for early interventions. In this dissertation, I use innovative methods and sources of data to guide decision making in AIS prevention efforts. First, anglers are major vectors of AIS and identifying pathways of angler movement provide an opportunity to intervene and stop the spread. Angler movement data has previously been restricted by costly angler surveys with low participation rates. Using catch logs from a widely popular angler app, I built an angler network that revealed an invasion superhighway that spanned the United States and likely contributed to the spread of two prolific invaders (e.g., Dreissena and Myriophyllum species). Next, identifying lakes that are vulnerable to invasion provide an opportunity for early detection and intervention. I used machine learning models to predict the presence of five aquatic invaders that have spread across the Upper Mississippi River basin. The models identified important features that make a lake vulnerable to invasion and identified high-risk sites for prioritizing AIS management. Then I discussed methods of dealing with poor data quality in big data used for invasive species management and proposed methods of data gathering to improve future studies. Lastly, targeted education and inspection of the most transient anglers would be valuable in slowing the spread of AIS. The variability in angler experience, skill, and motivation, means that anglers pose a heterogeneous risk to spreading AIS. In the last study, I used the angler app data to infer several characteristics of angler behavior and tendencies to predict transience. The results provided information to better inform AIS public outreach and enforcement campaigns. By identifying the pathways, locations, and people for targeted AIS management, this dissertation provides tangible results and scalable methods for preventative efforts. This work highlights the important role anglers can play in the prevention of AIS spread. However, it also demonstrates a need for better large-scale data management and quality improvements that are necessary for studying AIS at an appropriately large-scale.