Case Study: Biodiversity

Case study description:

A hypothetical state and federal government joint initiative is formulating a land management plan for “Region B”, which has a diverse regional climate including a gradient from the coast to mountain ranges. Rather than manage a particular flagship species, the initiative will manage overall biodiversity and ecosystem services both in terms of existing threats and potential new threats. In this context, the initiative partners would like to know the likelihood that climate change would lead to significant changes in bioclimatic envelopes, which would affect their choice of management settings. They would like a detailed analysis of the confidence behind the projected changes. The broad plan options are:

Policy

Rationale

Climate

No change

Analysis of changes due to all factors do not warrant any action

Projected change to the climate would have negligible impact on biodiversity

Manage non-climate threats

Factors (e.g. weeds and pests) are a far bigger threat now and in the future than changes to the climate

Projected change to the climate has an effect much smaller than other threats

Promotion of ecosystem resilience and monitor

Knowledge is not yet sufficient to dictate specific actions, so take general actions to the improve health and resilience of the system

Effect of changing climate is not yet certain enough to suggest specific action

Manage threats, improve connectivity

Managing existing threats must continue, but with an increase in management of habitat for biodiversity (e.g. wildlife corridors, preservation of refugia)

Climate change likely to induce significant changes to species distributions and greater dependence on refugia

Set time-frame and risk:

The initiative has the remit to manage to 2100. The study would like to consider all factors of uncertainty, so would like to consider all RCPs and model ranges using a “worst case”, “maximum consensus” and “best case” Climate Futures framework approach for each emissions scenario (RCP) for multiple timeframes (2050, 2070 and 2090).

Analysis context:

The relationship between Region B’s biodiversity and a changing climate is a major source of uncertainty. This is partly due to poor datasets of species distributions and a lack of longitudinal datasets. This limits informed decision-making. Therefore climate projections are in fact a secondary source of uncertainty. Despite this, very marked changes in climate could potentially still lead to confident conclusions about biodiversity change. For example, marked increases in temperature or reductions in water availability could have a strong influence on numerous species distributions. However, even if this is established, the best range of interventions may be difficult to establish and uptake by stakeholders may be limited.

Back of the envelope:

“Worst case” temperature increases and rainfall reductions under high emission scenarios by 2090 are +4 °C and -15% respectively, suggesting major action is warranted.

However, higher spatial resolution analysis is useful, especially when analysing biodiversity corridors and refugia.

Key context for choosing datasets:

The models of species distribution and biodiversity used require monthly climatological monthly averages, so the spatial dimension is very important but temporal variability is not relevant.

  1. Keep uncertainties in perspective - the greatest uncertainty may not be in the climate change data
  2. Sub-setting global climate models and regional climate projections is important to assess – range of results
  3. Scaling by the mean is ok – avoids problem of bias and difficulty of complex scaling
  4. Useful to account for seasonal cycle (e.g. use the ‘BIOCLIM’ variables)
  5. Downscaling may add value to the change signal for this region and this question, particularly with respect to rainfall change
  6. Analysis needs datasets with fine spatial resolution that accounts for fine-scale gradients such as altitude, slope and aspect

Choice of datasets:

Generate present climatological monthly averages of mean temperature and rainfall from historical data (e.g. from Bureau of Meteorology). Generate the equivalent data for the future from a “worst case”, “maximum consensus” and “best case” (using Climate Futures ) from the global climate models and relevant downscaling in each of the emissions scenarios for 2090 using mean scaling of fine-resolution observations. Once these surfaces are generated, do further scaling to very fine scale using appropriate software.