One of the ways in which climate change threatens biodiversity is through driving species’ range shifts. This is where species shift their distributions in response to changing climate conditions, to maintain their climatic niche. Typically, we would expect this to result in species shifting their ranges to higher latitudes and elevations. Yet, we often observe far more complex responses. This may be a result of the influence of land use change, persecution by humans, or other site-specific factors. Alternatively, species traits and their ability to adapt to change may affect how well that species can respond i.e. the dispersal ability of the species can affect if a species can track changes in climate. In this analysis we wanted to assess the extent to which climate change has driven recent range shifts relative to these other factors.

To do this we used data on the long-term changes in the breeding ranges of 378 species of European breeding birds. These data came from two Europe-wide bird distribution atlases  published 30 years apart, covering an area of 5,292,500 km2. The data for the first atlas (1985–1988) were collected at the onset of a period of major climatic change across Europe. The publication of the second atlas (data collected 2013–2017) provided us with a unique opportunity to quantify changes in species ranges at a continental scale over three decades of substantial environmental change. 

Using species distribution modelling (SDMs), we found that species’ observed range shifts over the 30 years did not match what would be expected based on changes in climate and land use. Specifically, we would expect that as temperatures increased across Europe, species would shift their ranges in a typically northward direction to maintain their climatic niches. What we found, however, was that the direction of observed species’ range shifts was far more variable. We also found that the rate of range shift was approximately 50% slower than would be expected given observed changes in climate alone.

We then explored the extent to which climate, land cover, and species’ traits drive local colonisation and extinction events: the two processes that determine species’ range shifts. We found that the dominant factors explaining the occurrence of colonisation and extinction events was how close an area was to nearby ‘source’ populations and the underlying suitability of climate in an area. The role of proximity to source populations reflects what we expect based on metapopulation theory and the ‘rescue effect’ . Less isolated areas more likely to receive dispersing individuals and therefore are either a) more likely to be colonised, or b) more likely to have a susceptible population rescued from extinction. This finding highlights the importance of maintaining networks of local populations to limit extinctions and to make populations more robust to the effects of climate change.

It is important to stress that our findings do not mean that climate change does not result in species range shifts. Rather, our findings suggest something more intriguing. In some areas ‘colonisation lags’ may result in species being unable to track improving climate, perhaps due to habitat or prey not yet being available in new sites. By contrast, fewer extinctions occurring in areas where we predict them to occur might be evidence of ‘extinction debts.’ Such debts occur when species are committed to eventual extinction due to unfavourable climate, but they nonetheless manage to persist, sometimes for lengthy periods, because key limiting factors, such as their preferred habitat, take some time to alter.

We also found a key role of non-climatic factors in driving local colonisation and extinction events. Specifically, we found that persecution still has a role in limiting European birds, highlighting that such things are still a major problem for many species. However, the rapid recovery of some species from past persecution or poisoning provides hope that populations can often rebound once such impacts are controlled.

 Our study highlights how complex species’ range responses to climate change can be, and how they can be moderated by other factors such as land use change, persecution, and species’ traits. If we are to robustly forecast how species may shift their ranges in response to future environmental change, we need to consider the roles of other complex process that can drive colonisation and extinction events.

 Finally, we would like to thank the European Breeding Bird Council, its partner organisations, and the thousands or ornithologist who contributed to the data collection. Without their hard work this work would not have been possible.