Risky Reserves

Planning for ecosystem services under sea-level rise using portfolio theory

By Rebecca Runting & Jonathan Rhodes (University of Queensland)


Key messages:

We used Modern Portfolio Theory to help manage risk and uncertainty in planning wetland reserves for climate change.

We compared spatial plans that resulted from our risk-sensitive approach to reserve selection that ignored risk to determine whether explicitly accounting for risk alters planning outcomes.

We demonstrated that incorporating sea-level rise, but ignoring uncertainty, is a high-risk strategy. In contrast, diversifying site selection through Modern Portfolio Theory can ensure the supply of ecosystem services by reducing the risk of failure across all sea-level rise scenarios.


When we think about climate change we often think about how hot or dry it is going to get, but don’t always make the connection to how this will affect all the benefits we get from the natural world, such as food, clothes and opportunities for recreation. Making this connection is crucial if we want to develop effective strategies for climate adaptation. But how do we do this given we are often really uncertain about how climate change will affect the benefits we get from nature? An inability to incorporate this uncertainty could mean our plans may fail to protect species and critical ecosystem services in the long-term.

We propose a method for incorporating this uncertainty into our planning. It involves extending a risk-sensitive approach to resource allocation from finance called Modern Portfolio Theory (MPT). The approach was developed by Nobel Prize-winning economist Harry Markowitz. It reduces risk by investing in combinations of (financial) assets that have negative correlations over future states of the world (or at least weak positive correlations). We’ve applied this approach to spatial reserve selection (Runting et al, 2018).

MPT has been applied to non-spatial problems in the management of species, populations, and ecosystem services. Some recent studies have also considered spatial planning units as ‘assets’ to allow for overall risk to be reduced by allocating conservation investment across space. However, these approaches cannot be directly applied to many planning problems that include discrete site selection, multiple objectives, and a consideration of connectivity.

We extended previous applications of MPT by incorporating these additional requirements into a risk-sensitive, spatially-explicit approach to reserve selection that maximises multiple conservation objectives, including ecosystem services, whilst hedging risk under climate change uncertainty and ensuring connectivity. This formulation closely resembles the types of problems conservation planners typically face, whilst accounting for risk in a mathematically rigorous way.

To demonstrate the value of our approach, we applied this approach to designing a reserve system under uncertain rates of sea level rise in Moreton Bay, Queensland. This area contains internationally important coastal wetlands and essential ecosystem services, while facing very high urban development pressures.

To design our reserve system, we first simulated hundreds of scenarios (804 to be exact) of how wetlands in Moreton Bay will change through to the year 2100. These scenarios incorporated uncertainties in future sea-level rise, elevation data, and other biophysical parameters using the Sea Level Affecting Marshes Model (SLAMM; see Decision Point #67).

We then optimised our risk-sensitive reserve design for three conservation objectives (Figure 1);

  • Wetland area (by hectare)
  • Blue carbon sequestration (Mg CO2 per year)
  • Nursery habitat for fisheries (by hectare)

For comparison, we also developed conservation plans based on the averages of each of the four IPCC projections of sea-level rise.

Our scenarios showed that as sea levels rise there would be substantial change in the distribution of wetlands by 2100. Mangroves would migrate landward, replacing saltmarsh, melaleuca, and dryland areasDPoint #106 (Oct 2018) (high res pdf for printing)_Page_10_Image_0001 (Figure 1). However, there was also considerable uncertainty surrounding these future distributions. Spatially, the highest uncertainties occurred at the lowest and highest elevations of the future wetland distribution due to the variation in potential losses (continual inundation) and gains (landward movement) in the coastal wetland extent. This variation in the future extent and type of coastal wetlands also affects the ecosystem services that flow from these wetlands, which exhibited even greater variation than the distribution of wetlands (Figure 2).

DPoint #106 (Oct 2018) (high res pdf for printing)_Page_11_Image_0002We found that it is possible to reduce risk but this comes at the expense of reduced levels of ecosystem services (Figure 3). However, approximately 50% of the risk could be reduced for only a 25% reduction in the level of services. We also found that incorporating sea-level rise while ignoring uncertainty is always a high-risk strategy, even when planning for worst-case scenario sea-level rise.

DPoint #106 (Oct 2018) (high res pdf for printing)_Page_11_Image_0001Accounting for uncertainty resulted in conservation planning solutions that provided multiple ecosystem services with a relatively low risk of failure across all climate scenarios. Reducing risk also changed the spatial configuration of the reserve network considerably, relative to planning solutions that ignored uncertainty. So, knowing how much risk you are willing to accept is important for determining which plans are likely to be best.

Accounting for risk improves the resilience of the reserve network through the diversification of sites and therefore helps ensure the continued supply of ecosystem services into the future. This method is likely to be of use in other spatial planning contexts, particularly for cases where the impacts of climate change on species, ecosystems, and their services vary spatially over different climate change scenarios.

Incorporating the impacts of climate change, and associated uncertainties, into spatial conservation planning, is critical to ensure the continued provision of valuable ecosystem services. Our investigation advances our understanding of how to spatially manage ecosystem services that are impacted by climate change, particularly where there are multiple objectives, uncertainties, and land use decisions to be made. In doing so we have provided tangible spatial solutions to manage our environment in an era of global change.

More info: Rebecca Runting r.runting@uq.edu.au

Reference: Runting RK, H Beyer, Y Dujardin, CE Lovelock, BA Bryan & JR Rhodes (2018). Reducing risk in reserve selection using Modern Portfolio Theory: coastal planning under sea-level rise. Journal of Applied Ecologyhttps://besjournals.onlinelibrary.wiley.com/doi/pdf/10.1111/1365-2664.13190

Leave a Reply