Designing landscapes that are robust to climate change

Background: To mitigate the threats of climate change and habitat loss, a species' future climatic range needs to contain enough suitable habitat that can be accessed from current habitat. Policy makers and conservation managers need models to help them design effective and connected habitat networks. These models need to predict the fate of populations in a very wide range of candidate landscapes, but current simulation models that faithfully capture spatially explicit individual-based population dynamics are too slow for this purpose. Alternative methods that can evaluate landscapes rapidly enough, including our own Condatis software, are not based on population dynamics and lack empirical validation.
Objectives: 1. Develop novel, computationally efficient, population and metapopulation models for persistence and range expansion on a spatially explicit habitat network. 2. Develop methods for rapidly evaluating the effect of landscape changes on persistence and range whose relationship to conservation targets are unclear; and there is scant empirical evidence that they recommend the best habitat configurations.
Timeliness: Policy makers and conservation practitioners need tools of this type now: our project will ensure that decisions made in landscape planning are supported by the best ecological knowledge. We will use a new mathematical framework (Cornell et al 2019) which computes predictions for spatially-explicit population models orders of magnitude more efficiently than simulation. We will also exploit the recent advances in JH's Condatis software (Hodgson et al, 2015), which show how to compute efficiently the effect of modifications in the landscape.
expansion 3. Empirically validate the models against observed range shifting in UK Lepidoptera species.
Novelty: This project will overcome the key weaknesses underlying current tools to support this sort of landscape-scale decision, specifically: they are not based on an underlying population dynamics model; they are based on metrics whose relationship to conservation targets are unclear; and there is scant empirical evidence that they recommend the best habitat configurations.