Climate change and constraint: why can’t more species respond to rapid warming?

Our ‘muse.’ A male Hudsonian godwit on the tundra in Churchill, Manitoba, Canada. Photo by Andrew S. Johnson.

By Nathan R. Senner @nrsenner

One of the fundamental problems currently facing scientists and conservationists is determining how best to identify the species that are most likely to succumb to climate change-related declines and extinctions. Despite the fact that there have been countless attempts at developing such a framework — focusing on everything from species’ life-history traits, to their range sizes, to the rate at which the regions they inhabit are warming — we do not seem to be any closer to achieving this goal than we were a decade ago.

A couple of years back, Maria and I were talking about another paper that we were writing together and how broadly its results might apply to other species. In that paper (Senner, Stager, and Sandercock 2017, Oikos), we studied two populations of a long-distance migratory bird, the Hudsonian godwit (Limosa haemastica). We found that one of the godwit populations was able to properly time its reproductive events in relation to the local resource phenology in spite of rapid, directional warming. However, we found that because of an asynchronous climate change regime — when a region is warming during some parts of the year, but cooling during others — the other population mistimed its reproductive events and suffered severely reduced reproductive success. During our discussion, we kept coming back to two questions: (1) How common are these asynchronous regimes? and, (2) How might their distribution affect the ability of individuals to disperse among different regions within their range?  

Thus the idea for our current paper was born. To fully address these questions, though, we needed to incorporate model simulations into our efforts to help determine how asynchronous regimes affected dispersal. We knew that these simulations would be strongest if we placed them within a population-genetic framework. Enter Zac. After a series of conversations with Zac, we arrived at a plan: We would use long-term climate data to identify how common asynchronous regimes currently are in North America and then use the results of those analyses to parameterize models of how the distribution of those regimes affect gene flow among populations.

What we found surprised us (Senner, Stager, and Cheviron 2017). Our results indicate that more than 70% of terrestrial North America is affected by asynchronous climate change regimes. Furthermore, the months that are most likely to be cooling are concentrated during the boreal spring, when many species are deciding when and where to breed, thus heightening the potential that they become mismatched with their local resource phenology. Finally, we found that the spatial distribution of asynchronous regimes is quite heterogeneous, with some regions experiencing many different climate change regimes within relatively small geographic areas. Thus, when we investigated how asynchronous regimes affect gene flow, we found that the current distribution of asynchronous regimes is likely a strong barrier to gene flow among populations within widespread species, especially when asynchronous regimes impose strong selection on the populations experiencing them.

While our results do not provide a new framework for predicting which species are most at-risk to the effects of climate change, they do suggest that the types of conditions that can impose constraints on a species’ response to climate change are common and widespread. Thus, although some species may be able to respond to rapid, directional warming, they may also face other constraints that exert a stronger influence on their adaptive potential. We therefore now need to focus our efforts on identifying those constraints.

Comments