Development and testing of an object-oriented model for adaptively managing human disturbance of least tern (Sternula antillarum) nesting habitat

Abstract

Botany Bay Plantation Wildlife Management Area is one of only four known natural nesting sites for least terns (Sternula antillarum) on the South Carolina coast, and more than 20,000 people visit the 3-mile stretch of beach each nesting season since it opened to the public in 2008. We developed a simulation model for the adaptive management of human disturbance on nesting habitat for least terns, which is described using the ODD (Overview, Design Concepts, and Details) protocol. The simulation uses the Questions and Decisions (QnD) system framework, an object-oriented modeling approach that emphasizes interactions among key ecological and social components. The model is designed to explore the potential results of management actions and identify key indicators for future monitoring. Initial model parameters were developed from existing literature, expert opinion and on-site ecological observations as few baseline data were available for the recently opened site. In addition, substantial uncertainty surrounded the nesting behavior of least terns related to productivity levels, responses to human disturbance, and overwash tides. In order to test model assumptions and performance, parameters are assigned stochastic probability distributions made explicit to wildlife managers and can be updated as additional monitoring data become available. Simulation results suggest that Botany Bay colonies are at significant risk of collapse. Monte Carlo simulations of the model (n=500) show that current management practices result in a median productivity of 0.08 fledglings per breeding adult, well below the minimum productivity threshold of 0.13 needed to encourage site fidelity. But closing one particular section of beach during the nesting season would result in a global median productivity of 0.26. A sensitivity analysis identified one variable in particular as important to monitor for reducing uncertainty in model results: the duration of tern defense response from human disturbance and the variation of this response among spatial units. These simulation results are intended to initiate an ongoing and adaptive monitoring program for improving colony productivity at the site.