How invaded is our lakescape?: Systematic surveys indicate biases in existing invasion records

Numbers of ecological invaders in North America are increasing exponentially, and their impacts on native ecosystems and economies are increasing.  These invasions are too numerous and widespread for researchers and managers to effectively monitor and manage at landscape scales, so citizen involvement particularly in the detection of new invaders is necessary.  For instance, there are over 2500 populations of aquatic invasive species across over 1100 lakes in Wisconsin, USA, alone.  In many datasets, citizen and volunteer-led efforts may form the majority of these detections.  These datasets also provide information for developing ecological niche and other habitat suitability models, predicting future invasions, and forming management plans.  However, since data collections are often not standardized, they may lead to biased assessments of invaded ecosystems and underrepresentation of the overall scale of invasion across a landscape.  In this study, we used a stratified random sample of over 450 lakes in Wisconsin to answer 2 simple questions: 1) what proportion of lakes are invaded? and 2) to what extent do citizen-derived data bias interpretations of landscape-wide patterns of invasion?

Lakes in our survey ranged from those completely isolated from boaters to large lakes with multiple boat launches, and from unsuitable to suitable.  For each class of lake, we quantified the proportion of lakes invaded in our survey and compared it to the proportion invaded in previous, existing datasets.  The proportions were multiplied by the overall abundance of each lake class across the state and summed to estimate the total number of invasions.  Overall, we estimate that about 42% of lakes in Wisconsin are invaded, including 100% of large, suitable lakes with public access.  In contrast, only 14% of these lakes were known to be invaded based on existing records.  Although invasion was less likely in smaller, less suitable, and less accessible lakes, these invasions tended to be less well represented in existing data.  Furthermore, many populations of relatively low-profile invaders such as snails were initially undetected.  In contrast, invasions of large lakes with high levels of human visitation and by high-profile "poster child" invasive species were well-represented in existing records.  These results demonstrate the extensive scale of invasion and highlight the needs for systematic surveys for accurate interpretation of landscape-scale invasion patterns.