A Dynamic Principal-Agent Model of Human-Mediated Aquatic Species Invasions

This paper presents an integrated ecological and economic model of aquatic plant species invasions in which resource users are the primary vector of species colonization. A random utility model of boater behavior is combined with ecological information about the invader and the landscape to form a dynamic principal-agent model in which the principal is a manager concerned with the interseasonal spread of invasive species across lakes, and agents are recreational boaters making a series of intraseasonal trip decisions to maximize random utility during the course of the season. Agent behavior is sensitive to both the degradation of environmental quality on colonized lakes and the actions taken by resource managers to control the spread of the invasive species. As such, this paper represents the first attempt in the literature to endogenize resource user behavior in the management of a species invasion, allowing us to appreciate the full impact of different policy scenarios. The results of a simple, simulated case study of an Eurasian watermilfoil (myriophyllum spicatum) invasion of a hypothetical system of four lakes are presented. These results highlight some interesting aspects of the optimal management policies under two different management objectives: maximizing welfare and minimizing milfoil spread. The size of the problem is a significant obstacle to a "real world" application. Recent advancements in both random utility modeling and dynamic programming simulation models could considerably ease these constraints.