Are avocados toast? A framework to analyze decision-making for emerging epidemics, applied to laurel wilt

CONTEXT: When emerging pathogens threaten global food security, collective action for disease management is key for protecting food systems. We evaluate how the informal exchange of information about epidemic and economic outcomes can influence the management decisions of individuals and the resulting epidemics, in the context of the avocado laurel wilt epidemic in south Florida. OBJECTIVES: In scenario analyses, we addressed how socioeconomic networks, laurel wilt epidemic networks, policy incentive structures, and social behaviors combine to influence (a) information exchange across this re-gion, (b) growers' decisions about disease management, and (c) epidemic spread and yield loss. We identified the scenarios in which regional avocado health fared best. METHODS: We built an agent-based model to simulate laurel wilt epidemic expansion and establishment across south Florida over a 10-year period. The model used parameters specific to patterns observed and quantified from the laurel wilt epidemic in south Florida. Based on the locations and sizes of avocado orchards there, we simulated disease expansion and information dissemination through multilayer socioeconomic and epidemic networks and evaluated the effects of "carrot" and "stick" policy incentive structures and behaviors like "stub-bornness" in decision making. Scenarios were simulated for multiple parameters across a 10-year time period, and the regional health of avocado and management decisions of growers were analyzed.RESULTS AND CONCLUSIONS: Increased social connections led to lower regional crop health due to increased exchange of information reinforcing selection of less expensive but less effective management choices. This in-formation exchange was particularly impactful during the lag phase of epidemic expansion, when the cost of disease management outweighed the cost of disease. Managers who were resistant or "stubborn" against adopting these less expensive and less effective management strategies, particularly during the lag phase of epidemic expansion, contributed to greater regional health. In these scenarios, growers responded more to policies which penalized individuals than to policies which rewarded individuals. SIGNIFICANCE: By quantifying varying degrees of stubbornness, and how growers may weight past experiences and new information, we represented key aspects of decision making and its many influences on regional col-lective action in this novel agent-based model. The model demonstrates the caveats of information exchange across social networks during epidemics, and the valuable role that policy makers and informed educators can have, particularly during the lag phase of epidemic expansion. Decision makers and stakeholders must under-stand the influences of information exchange to overcome the challenges of collective action for crop health.