By Quirin Schiermeier

Climate change is affecting the world in many ways. But attempts to directly link local changes in species distribution and biodiversity to climate warming hold little promise, ecologists warn in Nature Climate Change. First author Camille Parmesan, a population biologist at the University of Texas in Austin, explains why.

You argue that attempts to attribute the degree of local changes, for example declines in individual plant and animal species, specifically caused greenhouse warming are misguided. Why?

It is important to look at all the different things that could cause a decline, including climate change. But when it comes to managing and conserving species and ecosystems, trying to figure out exactly how much of any one particular decline is due to greenhouse gases is not necessarily helpful, and may actually not be possible.

You can, of course, attribute various individual biological changes to climate events, and even climate change, provided you have long-term studies. But linking observed changes to the man-made component of climate change requires a different scale. That level of attribution is best done for large areas -- the size of northern Europe or the western United States. The more local a scale you look at, the harder it is to link single events to greenhouse-gas-driven global climate change.

Take the endangered Quino checkerspot butterfly [Euphydryas editha quino] of Southern California. We do know that climate change is important: if you dry and warm the butterfly's habitat it will cause increased starvation and extinction. But many populations are also affected by an invasive geranium from the Mediterranean which is out-competing the butterfly's host plant. This is further aided by air pollution from Los Angeles and San Diego, as nitrogen fertilization helps the exotic geranium take over. It just doesn't make any sense to ask what percentage of the decline is due to anthropogenic climate change -- from a scientific standpoint it doesn't have much value. What you would be better off doing instead is manage for invasive geranium, lower the nitrogen pollution and set up new reserves that anticipate climate change -- that is, placed in areas the butterflies can colonize as climate shifts.

So how is climate change affecting Earth's flora and fauna?

Climate change is impacting biodiversity worldwide. Spring comes, on average, two weeks earlier. Almost two-thirds of species, including many birds, frogs, butterflies, trees and grassland flowers, breed or bloom earlier. More than 50% are changing where they live. There is a consistency in the global pattern of more than 1,700 species we are studying which tells us that the changes are linked to some common global force and are consistent with what you would expect from a warming world.

So we can say with confidence that, for example, the date of flowering in northern Europe has advanced by two weeks. But we cannot be sure that global warming is the reason a local butterfly or wild flower species is becoming extinct or expanding northwards?

Yes. Phenology signals are clearer than changes in species distribution. Plants and animals often respond to rising winter temperatures -- whether due to climate change or to increased urbanization, which also causes warming.

In a local area -- a given park or reserve -- conservation managers need to know what they can do to manage for a complex set of changes that may include habitat fragmentation, pollution, presence or absence of certain predators, and so on. It does partly depend on how climate sensitive a species is. But we shouldn't focus too much on how a species is impacted by climate change on its own, that's much too narrow a focus. Other factors are often far more important and the ultimate impact of climate change on any given population is going to be dependent upon how stressed that population is by these other problems.

Scientists have previously linked the extinction of Costa Rica's iconic mountaintop golden toad [Bufo periglenes] to climate change. Was that premature?

No -- I believe that was a correct assessment. The golden toad was endemic to Costa Rica's Monte Verde preserve. It went through three major population declines -- each preceded by an extremely dry and hot year. The third such event was followed by toad extinction.

What has happened looks like a clear case of extinction driven by three extreme years. So yes, climate did cause the extinction, but we have only 17 years of data. To say whether the extreme dryness was caused by rising greenhouse gases requires looking at the region as a whole and asking whether -- or by how much -- such events are more likely under a greenhouse-gas-driven global climate. This may appear a subtle distinction, but it is this distinction that we are trying to express.

Species distributions are also shifting. The long-spined sea urchin [Centrostephanus rodgersii], for example, has moved from the warming seas off mainland Australia and has invaded more temperate waters off Tasmania. Likewise, many terrestrial plants and insects are moving or expanding pole-wards. Are these not clear fingerprints of climate change?

Sure, the sea urchin is probably shifting due to warming waters. As it shifts, it's been devastating local ecosystems. But it may not have become invasive if humans had not already over-fished the rock lobster off Tasmania -- we'll never know. So there's an interaction between climate change and overfishing. Same with coral reefs: the fact that many are dying after high ocean temperature events may have something to do with humans stressing them with pollution, dynamite fishing, recreational activities and coastal development. Death rates may not have been so dire if the reefs were not overstressed by other things humans are doing -- again, we'll never know.

Some people want to deconstruct any observed biological change into separate causes and put percentage figures on them. In my opinion that's not worth spending the time on. Because in real systems, all these things interact, we just can't know exactly how much of a local change is due to greenhouse warming.

If you have data over a large area -- like our butterfly study of all of Europe -- then you can definitely say the northward shifts of two-thirds of European butterflies in the UK, France, Sweden, Finland and Estonia are linked to long-term climate warming in Europe. But if you go to one population in the UK and see that, at a single site, there have been new populations formed further north, you will have a harder time making the greenhouse-gas link and you probably just shouldn't try.

The Intergovernmental Panel on Climate Change (IPCC) supports an ever-more-detailed approach to biological attribution -- not least to inform conservation efforts. Do you think there is a better way to approach conservation?

The IPCC exists to distil science to answer policy makers' questions. But it is perhaps too often influenced by policy makers' black-and-white view of the world. We need to train policy makers to think of probabilities and likelihoods and interactions. They don't like it -- but this is the accurate way to describe the science. There are times when it is appropriate to tell people that these are not the right questions -- this is one of those times.

Does your critique include physical attribution studies -- attempts, that is, to pin down how greenhouse warming affects the likelihood of specific weather-related events, such as heatwaves, heavy rain and floods?

No, we have very different approaches. If climate scientists can get to where they feel comfortable in terms of assigning attribution on very local scales, fine. We would even benefit from it. But biology and ecology are fundamentally different from -- and I argue more complex than -- climate science -- humans are doing much more harm to wild species than just adding carbon dioxide to the atmosphere. How much wildlife is harmed by global warming depends on how stressed the system is by all the other things humans are doing locally.