3 minute read

Let's start with a doomed species. The white lemuroid possum is a spectacular creature that lives in the high altitude rainforest of Australia. As climate change progresses, the cool conditions this possum thrives on will retreat uphill, and eventually disappear. And with the demise of these cool conditions, so too this astonishing creature. Already, scientists are documenting precipitous declines in this possum. It may well be gone within a decade.

Climate change is underway, and it's bad for biodiversity. Of late, biologists have gone from demonstrating the impact of climate change on organisms -- changes in species ranges, changes in breeding times and so on -- to wondering how we might mitigate this impact.

Halting climate change would be the best strategy, of course, but this doesn't look likely to happen soon. Instead, we are looking at having to move organisms to places where the climate will be suitable for them. The mountains the white lemuroids call home are getting too hot, but are there other mountains further south (or taller) that remain cool enough even as the planet warms?

In principle, this is a simple recipe:

Step 1, we look at where a species currently lives;
Step 2, we dial up our best predictions for the future climate, and;
Step 3, we look for places where the future climate matches the current climate where the species lives.

We are engaged in picking climate pairs.

Simple, right? Well, kind of. When you think on it a little deeper you come to a problem. Climate doesn't have simple units. You can't measure climate in centimetres, or degrees celsius. In fact, climate is multidimensional.

This multidimensionality is an irritating fact. It is ludicrously simple to compare several locations for how well they match for temperature. It is also simple to compare locations for how well they match for rainfall. But it is not straightforward to compare locations for how well they match for both temperature and rainfall. Which of these sites -- B or C -- is best matched to site A?

Site | Temperature (degrees) | Rainfall (mm)
A | 25 | 1000
B | 27 | 1000
C | 25 | 1200

Why is this problem difficult? Because rainfall and temperature have different units. And let's remember that temperature and rainfall are just two aspects of climate. There are many others.

The solution to the multidimensionality problem is intuitive, however. If temperature is more important to our animal, then we give that 2 degrees difference greater weighting than the 200mm of rainfall (and so we choose site C). But this solution just raises another problem: how do we know that temperature is more important for this species? And how much more important is it, anyway?

Thanks for following me this far down the rabbit hole.

Now we are faced with the issue of working out the relative importance of each aspect of climate. In technical terms, we need to work out how to scale each of these aspects of climate against each other. To make this scaling we need a common currency: the equivalent of the US Dollar, or Carbon 12. If we convert each aspect of climate to this common currency, then it is easy to see which aspects are worth more than others. In biology, that common currency is fitness.

Converting temperature (or rainfall) to fitness is tricky, however. So we* thought to sidestep that problem and, instead, measure how strongly a species is adapting to each aspect of climate. The rationale being that if, across a species range, it shows strong local adaptation to temperature, but not rainfall, then we have good evidence that temperature is important to fitness, but that rainfall is, perhaps, less so. This "strength of local adaptation score"** gives us an objective way to scale aspects of climate, solve the multidimensionality problem, and get to picking climate pairs in the most effective way possible.

This capacity to pick sensible pairs is important. Not just for the white lemuroid possum, but for all species which we might consider the movement of individuals a useful strategy for mitigating climate change impact.

To read more, see our recent paper in Ecology Letters.

*we being Stewart Macdonald, John Llewleyn, Craig Moritz and myself

** this score will be the subject of another blog post at some point. It's interesting in its own right.