|HOFFMAN, KEVIN - California Department Of Food And Agriculture
Submitted to: Journal of Pest Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/18/2013
Publication Date: 3/20/2014
Citation: Manoukis, N., Hoffman, K. 2014. An agent-based simulation of extirpation of Ceratitis capitata applied to invasions in California. Journal of Pest Science. 87:39-51.
Interpretive Summary: This paper describes a method (using Agent-Based Simulation) for estimating the time to local elimination (extirpation) of invasive insects that differs significantly from the most common current approach, often called the “degree day” method. Under degree-day, average temperatures in the location of an invasive insect find are used to predict the thermal unit accumulation insects will experience, then this information together with historical average temperatures is used to estimate how long development should take for the invader in this area. An expert-determined number of generations is used to estimate how long quarantine should be enforced for the area of the find. The degree-day method does not consider several important real-world variables that will determine how long insects might persist in the area. The agent-based method used here 1)is demographically explicit, meaning that effects of small population sizes are included; 2)includes the possibility of rare events, like unusually low developmental rates or high mortality, and 3) includes consideration of human-induced mortality as a consequence of countermeasures taken when an invasive insect find is made. We have used the Agent-based method to study the time to local extinction of invading Medfly in California.
Technical Abstract: We describe and validate an Agent-Based Simulation(ABS) of invasive insects and use it to investigate the time to extirpation of Ceratitis capitata using data from seven outbreaks that occurred in California from 2008-2010. Results are compared with the length of intervention and quarantine imposed by the State, which was effective at locally eliminating C.capitata. In each case there is good agreement between the length of intervention and the time to extirpation indicated by the ABS, but the simulation suggests that the margin of safety is not the same in all cases. We also examined changes in the number of individuals over time in the ABS and conducted a sensitivity analysis for one of the outbreaks to explore the role of input parameters on simulation outcomes. While our implementation and use of the ABS in this work is motivated by C.capitata and takes extirpation as a postulate, the simulation is very flexible and can be used to study a variety of questions on the invasion biology of insects. The agents can also be extended to include more complex behavior and to consider factors such as spatial location in a landscape. Any use of this ABS or extensions of it will require careful parametrization and significant computing resources. These issues and approaches to applying the ABS to other questions are addressed.