Hide and seek in a warming climate

How are different animal species adapting to a warmer future?

Dr Mick Ashcroft

Dr Mick Ashcroft
Photographer: John Gollan © Australian Museum

Much of the world’s biodiversity faces an uncertain future as global warming continues across the Earth’s surface.

The scale and pace of the threat posed by a warming climate has alarmed many scientists, leading them to search for areas that may be least affected by having unusual or stable climates. Scientists reason that such areas could provide somewhere for species to expand and contract their populations over long periods or offer shelter from more extreme conditions. They have been termed ‘microrefugia’ but are also known as ‘cryptic refugia’ because of the difficulty in observing them within a landscape.


Before we can identify refugia, we first need to define them in an ecologically meaningful way. Scientists have known for years that refugia exist as patches in variable (heterogeneous) landscapes, but until recently our definitions of microrefugia have lacked precision or were poorly characterised. To address these shortcomings, we have been conducting detailed and intensive field studies over the last three years monitoring the microclimate at a range of sites in the greater Hunter region west of Newcastle.


Every six months, we set off from Sydney to collect data from a total of 150 climate loggers that automatically record hourly temperature and humidity readings. Each field trip takes around a week. The small, button-sized climate loggers have been placed in a variety of sites including rainforest gullies; steep slopes having different aspects; exposed rocky escarpments and even sand dunes. So far, we’ve amassed over six million observations.
Combining these data with detailed information about elevation and the local topographic features that influence temperature has produced a technique that can be widely used to quantify and locate potential microrefugia in a landscape. In short, the technique captures extreme conditions (either very hot or very cool), the degree of stability in these conditions, and how distinct these conditions are from the climatic conditions in the surrounding landscape.

To test and validate the technique, we successfully used it to predict the location of known communities considered to occupy refugia, such as rainforests that have progressively contracted in distribution over the last 2.5 million years, and alpine grasslands that have contracted over the last 15,000 years. So we know the technique works, but can it be applied more widely?


The vast scale of Australian landscapes complicates the incorporation of microrefugia into adaptation plans for future climate change. Indeed, some Australian land management agencies are responsible for land areas larger than certain European countries. This is where computer modelling plays a valuable role. In the past, complex global climate models (GCMs) have been used to identify refugia. A limitation is that many GCMs are based on cells measuring tens to hundreds of kilometres. We now know that temperatures vary at scales of less than one kilometre. In effect, these coarse-scale models can define features that are more correctly termed ‘macrorefugia’, and they certainly have a role in planning for climate change effects.

But another shortcoming is that GCMs fail to accurately estimate surface climate conditions because they are too coarse to take into account the very terrain features that help to decouple upper atmospheric conditions from boundary layer effects. This is a significant weakness because it is the surface climate that most organisms experience and that directly affects their survival: germinating seeds, tender saplings and the majority of terrestrial invertebrates and ground-dwelling vertebrates – all are at the mercy of near-surface temperatures.

Further, GCMs are based on temperature, more specifically ‘average temperature’. This measure has long been held to be important in understanding the distributions of species and in turn their responses to future changes in climate.

Our studies, however, show that it is the extremes and the intervening periods of stability that are more significant in a biologically or ecological sense. We hope that our work will lead to a paradigm shift in which the roles of temperature stability and isolation, rather than long-term averages, are properly considered in understanding ecological communities.

To the future

To test our hypotheses about the roles of temperature extremes and stability, our work at the Museum continues under an ARC-funded project partnered with the University of Technology, Sydney, University of NSW, NSW Office of Water and the Central West Catchment Management Authority. With PhD students, we are planning to conduct intensive biodiversity surveys across the full range of conditions found in the NSW Central West and supplement it with data from collections databases held in museums. The study will include surveys of deep rainforest gullies, exposed mountain ridges and steep rocky slopes.

Some of the questions to be answered include: are microrefugia adequately represented in our reserve network? does climate stability act to stabilise a community? and, are microrefugia more likely to support endemic species? If we are to meet the very real threat to biodiversity posed by climate change, we must work closely with land management authorities to identify microrefugia – not all of which will be represented within existing conservation reserves.

We might not be able to stop the inexorable creep of global warming, but perhaps we can help slow the loss of species through better management of refugia.

Dr John Gollan Research Associate & Dr Mick Ashcroft Spatial Analyst

Further reading

MB Ashcroft, JR Gollan, DI Warton & D Ramp, 2012. A novel approach to quantify and locate potential microrefugia using topoclimate, climate stability, and isolation from the matrix. Global Change Biology, online 2 March 2012. doi: 10.1111/j.1365- 2486.2012.02661.x 

Brendan Atkins , Publications Coordinator
Last Updated: