Model-based scenario planning to develop climate change adaptation strategies for rare plant populations in grassland reserves

摘要

Incorporating climate change into conservation decision-making at site and population scales is challenging due to uncertainties associated with localized climate change impacts and population responses to multiple interacting impacts and adaptation strategies. We explore the use of spatially explicit population models to facilitate scenario analysis, a conservation planning approach for situations of high uncertainty. We developed dynamic, linked habitat suitability and metapopulation models using RAMAS GIS to consider management and monitoring options for a grassland reserve in Minnesota (USA) in order to support a hydrologically sensitive rare orchid (Cypripedium candidum). We evaluated 54 future scenarios combining changes in drought frequency, increased depth to water table, and multiple configurations of increased invasive species cover and management. Simulation results allowed us to prioritize adaptation strategies and monitoring guidelines to inform adaptive management for our model system. For example, preventing further spread of invasive species into the current C. candidum population is an important low-risk resilience strategy for this site. However, under more serious climate change scenarios, higher-risk strategies, such as protecting critical recharge areas, become essential. Additionally, allocating limited monitoring resources toward detecting changes in depth to water table and assessing C. candidum population responses to severe drought will more efficiently inform decisions about when to shift from low-risk resilience approaches to higher-risk resistance and facilitation strategies. Applying this scenario based modeling approach to other high-priority populations will enable conservation decision-makers to develop sound, cost-effective, site-specific management and monitoring protocols despite the uncertainties of climate change. (C) 2015 Published by Elsevier Ltd.