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.

Policymakers and donors often need to identify the locations where technologies are most likely to have important effects, to increase the benefits from agricultural development or extension efforts. Higher-quality information may help to target the high-benefit locations, but often actions are needed with limited information. The value of information (VOI) in this context is formalized by evaluating the results of decision making guided by a set of specific information compared with the results of acting without considering that information. We present a framework for management performance mapping that includes evaluating the VOI for decision making about geographic priorities in regional intervention strategies, in case studies of Andean and Kenyan potato seed systems. We illustrate the use of recursive partitioning, XGBoost, and Bayesian network models to characterize the relationships among seed health and yield responses and environmental and management predictors used in studies of seed degeneration. These analyses address the expected performance of an intervention based on geographic predictor variables. In the Andean example, positive selection of seed from asymptomatic plants was more effective at high altitudes in Ecuador. In the Kenyan example, there was the potential to target locations with higher technology adoption rates and with higher potato cropland connectivity, i.e., a likely more important role in regional epidemics. Targeting training to high management performance areas would often provide more benefits than would random selection of target areas. We illustrate how assessing the VOI can contribute to targeted development programs and support a culture of continuous improvement for interventions.[Formula: see text] Copyright © 2022 The Author(s). This is an open access article distributed under the CC BY 4.0 International license.

Policymakers and donors often need to identify the locations and settings where technologies are most likely to have important effects, to increase the benefits from agricultural development or extension efforts. Higher quality information may help to target the high-payoff locations. The value of information (VOI) in this context is formalized by evaluating the results of decision making guided by a set of information compared to the results of acting without taking the information into account. We present a framework for management performance mapping that includes evaluating the VOI for decision making about geographic priorities in regional intervention strategies, in case studies of Andean and Kenyan potato seed systems. We illustrate use of Bayesian network models and recursive partitioning to characterize the relationship between seed health and yield responses and environmental and management predictors used in studies of seed degeneration. These analyses address the expected performance of an intervention based on geographic predictor variables. In the Andean example, positive selection of seed from asymptomatic plants was more effective at high altitudes in Ecuador. In the Kenyan example, there was the potential to target locations with higher technology adoption rates and with higher potato cropland connectivity, i.e., a likely more important role in regional epidemics. Targeting training to high performance areas would often provide more benefits than would random selection of target areas. We illustrate how assessing the VOI can help inform targeted development programs and support a culture of continuous improvement for interventions. ### Competing Interest Statement The authors have declared no competing interest.

Effective altruism is an ethical framework for identifying the greatest potential benefits from investments. Here, we apply effective altruism concepts to maximize research benefits through identification of priority stakeholders, pathosystems, and research questions and technologies. Priority stakeholders for research benefits may include smallholder farmers who have not yet attained the minimal standards set out by the United Nations Sustainable Development Goals; these farmers would often have the most to gain from better crop disease management, if their management problems are tractable. In wildlands, prioritization has been based on the risk of extirpating keystone species, protecting ecosystem services, and preserving wild resources of importance to vulnerable people. Pathosystems may be prioritized based on yield and quality loss, and also factors such as whether other researchers would be unlikely to replace the research efforts if efforts were withdrawn, such as in the case of orphan crops and orphan pathosystems. Research products that help build sustainable and resilient systems can be particularly beneficial. The "value of information" from research can be evaluated in epidemic networks and landscapes, to identify priority locations for both benefits to individuals and to constrain regional epidemics. As decision-making becomes more consolidated and more networked in digital agricultural systems, the range of ethical considerations expands. Low-likelihood but high-damage scenarios such as generalist doomsday pathogens may be research priorities because of the extreme potential cost. Regional microbiomes constitute a commons, and avoiding the "tragedy of the microbiome commons" may depend on shifting research products from "common pool goods" to "public goods" or other categories. We provide suggestions for how individual researchers and funders may make altruism-driven research more effective.[Formula: see text] Copyright © 2020 The Author(s). This is an open access article distributed under the CC BY 4.0 International license.

The geographic pattern of cropland is an important risk factor for invasion and saturation by crop-specific pathogens and arthropods. Understanding cropland networks supports smart pest sampling and mitigation strategies. We evaluate global networks of cropland connectivity for key vegetatively propagated crops (banana and plantain, cassava, potato, sweet potato, and yam) important for food security in the tropics. For each crop, potential movement between geographic location pairs was evaluated using a gravity model, with associated uncertainty quantification. The highly linked hub and bridge locations in cropland connectivity risk maps are likely priorities for surveillance and management, and for tracing intraregion movement of pathogens and pests. Important locations are identified beyond those locations that simply have high crop density. Cropland connectivity risk maps provide a new risk component for integration with other factors-such as climatic suitability, genetic resistance, and global trade routes-to inform pest risk assessment and mitigation.