Australasian Science: Australia's authority on science since 1938

Climate Extremes Matter Most for Biodiversity


How many species will succumb as temperature increases by more than 1°C, more than 2°C, and eventually to a predicted 4–6°C before the end of the century?

By Ary Hoffmann

The world is expected to warm by up to 4°C by 2070, but it is the extreme weather events associated with climate change that threaten biodiversity the most.

One of the main challenges facing biologists today is to work out when biodiversity around the world will start to fall over in a rapidly changing climate. Now that climate change due to anthropogenic causes has become accepted as inevitable and irreversible in the foreseeable future, the challenge is to understand when populations and species will start to decline and then become extinct within an area.

Will it occur within the next decade, or will it take much longer? What fraction of species will succumb as temperature increases by more than 1°C, more than 2°C, and eventually to a predicted 4–6°C before the end of the century? Will these declines then drive changes in entire natural communities of plants, animals and microbes, or is a process of substitution or evolution possible so that some ecosystems continue to persist in a form similar to what they look like today?

Answers to these questions are not merely of academic interest. Functioning ecosystems help to maintain water supplies for irrigation and human consumption, add fertility to soils, provide pollinators and natural pest control to allow for crop production, and provide buffers to detoxify human waste and store carbon. Any deterioration of ecosystem health has massive repercussions for our lives, and it is one of the reasons why ecologists can get frustrated by an unwillingness of the public and/or government to engage in the broader implications of climate change.

There are numerous signs that climate change is already causing large shifts in natural communities. These include shifts in the distribution of species along climate gradients such as the east coast of Australia, where some species of butterflies and other insects as well as birds are now being found further south. These organisms, as well as some mammals and reptiles, are also moving to higher elevations as conditions become warmer.

There are now detectable changes in the timing of reproduction by plants and animals in both natural and agricultural ecosystems. A well-documented local example involves bunch maturation and vintage in vineyards, which has shifted several weeks earlier in the past few decades even though there has so far only been limited warming of around 0.8°C.

To date, only a few species extinctions have been attributed to climate change, but a number of models predict that extinctions will start to become more common as warming effects build up. Biologists need to develop accurate ways of predicting these events so that steps might be taken to identify natural communities that are likely to be the most threatened, and to devise management strategies that might allow constituent species to move into new areas. It might also be possible to re­invigorate existing communities by moving genes and species around the landscape. As environments deteriorate, it may also become necessary to rescue iconic species by moving them well away from areas that they currently occupy.

Evaluating Threats

The traditional way of viewing threats faced by a species is to think in terms of the ability of individuals to survive and maintain reproduction under a range of conditions, such as a range of temperatures or humidities. Biologists then use this curve to define the maximum temperatures (CTmax) and minimum temperatures (CTmin) required to measure aspects of their life, such as reproduction or activity levels needed to find mates and escape predators. These measures of an organism’s fitness typically fall away more sharply at high temperatures than under low temperature conditions, reflecting the fact that the fitness of species is affected little until a thermal threshold is passed.

To measure the width and shape of these curves, biologists have typically carried out experiments across a range of constant temperature conditions. This might involve holding animals for a period at a predefined temperature and then measuring their activity. For short-lived organisms it may be possible to follow their entire life cycle at a particular temperature to see if they persist and reproduce across a generation. The example in Figure 1 shows the typical curve for a species of Drosophila fly, a small insect.

However, in natural environments and particularly terrestrial environments, organisms are typically exposed to highly variable conditions. This variability is especially pronounced when heat waves, cold snaps, floods or extended droughts result in periods of exposure to unusually extreme conditions.

Unfortunately, these types of conditions are expected to increase in frequency and severity under climate change. Evidence for such changes is captured in a report by the Climate Council (, and a recent example is the heatwave experienced by much of Victoria early in 2014 (Fig. 2). In this particular event, extremes were more than 15°C higher than average monthly temperatures.

Therefore, rather than being exposed to a range of constant conditions over their entire life cycle, organisms have to deal with a high degree of variability, which in turn can lead to complete failure of reproduction and in extreme cases death.

Acclimation to Extremes

There is an ongoing and active debate about the ways in which extreme conditions should be imposed on organisms in such a way that they are biologically meaningful. One approach is to source individuals from an environment and then expose them to increasingly higher or lower temperatures to determine when they reach some limit that is measured by a loss of activity, coordination or even death.

For instance, lizards are typically scored for their “righting” response, where they are placed on their back and their ability to flip over the right way is measured. The CTmax and CTmin of these short events fall well outside those where continuous culture is present.

However, this approach ignores acclimation – the extent to which the response to extremes is ameliorated by conditions immediately preceding those leading to a stress.

Acclimation is exhibited to some extent by all animals and plants – and bacteria for that matter. Cells contain essential machinery that helps them deal with stressful conditions, such as the production of proteins that can bind to and prevent the degradation of complex macromolecules.

Acclimation can be highly effective and increase the CTmax or CTmin of an organism by several degrees. The extent to which it helps depends on how fast the stressful extreme conditions develop, the conditions experienced just prior to the stress, and even the conditions encountered by the previous generation.

For instance, the heat and drought resistance of an insect or a fish can depend on the temperature that its mother experienced in the previous generation, as well as the conditions it experienced during development. As a dramatic example of this effect, quite a few species of Australian invertebrates survive hot and dry conditions in summer as diapausing eggs produced by mothers in response to cues that mark the end of spring.

As well as depending on acclimation, resistance to stress also depends on the rate at which stressful conditions are encountered. A gradual increase in temperature can be much more harmful than a sudden increase, most likely because damage occurs to individuals even before temperatures reach harmful levels. The way the effects of extremes are measured therefore needs to be relevant to natural conditions, but this is challenging particularly when species can locate small microclimates that reduce their levels of exposure, such as lizards seeking shade among rocks and small mammals moving underground.

Damage caused by stressful extremes may be cumulative over time, particularly when it cannot be adequately repaired. For instance, during heat waves across successive days, the levels of resistance of aphids decrease not so much because of the extreme conditions during the day but because night temperatures are also higher. Warm nights are not directly harmful, but they can prevent the insects from repairing damage before the next onset of daytime stress, decreasing their levels of resistance.

All of these factors highlight the challenges of measuring the way species are likely to deal with stressful extremes. Only by careful experimentation can we hope to collect the data needed to make useful predictions, yet very little is known about the extent to which Australia’s plants and animals respond to extreme conditions.

Extremes and Distributions

The importance of extreme conditions has recently been highlighted by comprehensive data we have collected on Australian fly species belonging to the genus Drosophila. This group of flies represents an excellent model system for investigating climate effects because different species are distributed along the length of the eastern coast as well as further inland, providing natural gradients of climate from the tropical north to the temperate south and dry inland areas. Some species are confined to damp northern rainforests while others have colonised cooler and/or drier areas.

By characterising the response of different species to constant temperatures, we have shown that the distribution of these species is limited not by the average conditions they experience but by short periods of extremes. Tropical and widespread species have remarkably similar fitness when cultured in conditions between 14°C and 30°C, but they differ in their response to both cold and warm extremes that extend well outside this range.

Based on these conditions and responses, it is possible to develop models that predict their distributions. The example in Figure 3 represents predictions for a widespread species and a narrowly distributed species, and occurrence points match modelled distributions tightly.

We can then use this information to predict what will happen to the species in the future. If temperatures rise by 2–4°C, as they are currently predicted to by 2070, what will the distribution of the species look like?

Unfortunately, the predictions are dire. Extremes are expected to exceed the conditions that can be tolerated by these species, resulting in the likely extinction of many species. Insects appear incapable of surviving hot conditions that extend a few degrees above those encountered currently. It is likely that dry extremes will also be poorly tolerated by many species. Thus while some species that are relatively robust in being able to persist in inland areas are expected to flourish under these conditions, it appears that much biodiversity is threatened by extremes. These predictions may well be conservative because they do not take into account the reduced level of heat resistance expected under night warming.

It’s possible that some species will escape these stressful conditions through genetic or epigenetic changes. Rapid adaptation through evolution has been demonstrated in some species of insects, plants, birds and other organisms in response to recent climate change, particularly in weedy and widespread species with short generation times where natural selection can be highly effective. However it is still not clear whether these evolutionary changes will be sufficient to keep up with climate change, and this is unlikely in species that already have restricted distributions and that lack genetic variation because of their small population sizes.

While human populations might persist in air conditioned comfort under heatwaves, our biodiversity has a far more limited set of options for adapting to extremes. In the end, we will all be affected because our health is closely tied to biodiversity in the natural environment.

Dealing with climate change is not about choices but about necessity, and preserving biodiversity is a requirement, not an option. Only mitigation of climate change effects through a concerted local and international effort is likely to have much impact on biodiversity in the long run by limiting extremes to increases that can be handled by most species.

Ary Hoffmann is Australian Laureate Fellow at the University of Melbourne’s departments of Genetics and Zoology.