Boats, Pathways, and Aquatic Biological Invasions: Estimating Dispersal Potential with Gravity Models

Abstract

Biological invaders can have dramatic effects on the environment and the economy. To most effectively manage these invaders, we should consider entire pathways, because multiple species are dispersed through the same vectors. In this paper, we use production-constrained gravity models to describe movement of recreational boaters between lakes – potentially the most important pathway of overland dispersal for many aquatic organisms. These models are advantageous because they require relatively easily acquired data, hence are relatively easy to build. We compare linear and non-linear gravity models and show that, despite their simplicity, they are able to capture important characteristics of the recreational boater pathway. To assess our model, we compared observed data based on creel surveys and mailed surveys of recreation boaters to the model output. Specifically, we evaluate four metrics of pathway characteristics: boater traffic to individual lakes, distances traveled to reach these lakes, Great Lakes usage and movement from the Great Lakes to inland waters. These factors will influence the propagule pressure (hence the probability of establishment of invasive populations) and the rate of spread across a landscape. The Great Lakes are of particular importance because they are a major entry point of non-indigenous species from other continents, hence will act as the origin for further spread across states. The non-linear model had the best fit between model output and empirical observations with r ² =0.80, r ² =0.35, r ² =0.57, and r ² =0.36 for the four metrics, respectively. For the distances traveled to individual lakes, r ² improved from 0.35 to 0.76 after removal of an outlier. Our results suggest that we were able to capture distances traveled to most but not all lakes. Thus, we demonstrate that production-constrained gravity models will be generally useful for modeling invasion pathways between non-contiguous locations.