This blog post is provided by Paul Acker, Francis Daunt, Sarah Burthe, Mark Newell, & Jane Reid and tells the #StoryBehindThePaper for the article “Strong survival selection on seasonal migration versus residence induced by extreme climatic events“, which was recently published in Journal of Animal Ecology.

Ongoing global changes are causing rapid and substantial alterations of environments to which species had previously adapted. Resulting mismatches between individuals and their environments can reduce survival and reproductive success, and hence threaten population persistence. In particular, climate change is causing increasing frequencies and intensities of extreme events, such as storms, heat waves, or floods. Such events and their environmental consequences are likely to be key drivers of spatio-temporal dynamics of biodiversity.

Most obviously, extreme climatic events can lead to mass mortality and reproductive failure, and thus directly increase risks of population extinction. However, individuals who manage to survive through these events may express characteristics that confer the ability to overcome or avoid the most adverse weather conditions. In other words, extreme climatic events may cause strong natural selection for characteristics that mitigate their negative effects.

If survivors keep expressing the same beneficial characteristics, the selected population might then be more likely to overcome further extreme climatic events in the short-term. If these characteristics are inherited by offspring, their frequency in the population will increase in the longer term, representing adaptive evolution towards greater resilience against extreme climatic events. The population could hence be ‘rescued’ from likely extinction via natural selection. Extreme climatic events thus offer valuable opportunities for evolutionary ecologists to observe selection and its evolutionary implications in action.

Seasonal migration is one key trait, which is widespread in animals, that could act in this way. Individuals that move to spend the winter in sufficiently distant or sheltered locations might escape extreme winter storms that affect their breeding areas, while individuals that stay resident in their breeding areas are severely impacted. However, the degree to which extreme storms do indeed cause strong selection for seasonal migration, and hence could potentially facilitate population persistence by rapidly changing how individuals are spatially distributed, had not been rigorously quantified. To tackle this issue, we focused on a partially-migratory population of European shags (Phalacrocorax aristotelis) breeding on the Isle of May National Nature Reserve in Scotland, which is known to regularly  experience winter storms leading to major ‘wrecks’ (mass occurrence of dead birds).

The physiology and ecology of shags make them highly vulnerable to the combined effects of cold, wet, wind, and rough seas that characterise winter storms and impede foraging and thermoregulation. To test whether shags can avoid such a lethal cocktail by migrating, we collected individual-based data on non-breeding season location and survival over multiple years. To achieve this, we marked individuals on the Isle of May with uniquely inscribed colour rings that are readable from up to 150m with a telescope. Then, during each winter since 2009, we visited the Isle of May and other locations across the east coasts of Scotland and northern England to look for ringed shags and determine which individuals were still alive and either resident (on the Isle of May and surroundings), or migrant (at other locations). This required an immense effort from dozens of fieldworkers and ‘citizen scientists’, who collected 43,214 sightings of over 2274 different individual shags that were known to breed on the Isle of May. While you are reading this blog right now, some of them might be out collecting the precious re-sightings that fuel our endeavour to understand ecological and evolutionary dynamics involving seasonal migration.

Some hardy observers were out when historically extreme storms occurred in late winters of the biological years 2012-13, 2013-14, and 2017-18. In 2012-13, coastal sea conditions in eastern Scotland were so rough that huge waves caused extensive damage to coastal infrastructure. In 2017-18, the storm caused by the anticyclone ‘Hartmut’ brought an extreme cold wave that was dubbed “the Beast from the East” by UK medias. These events caused major ‘wrecks’, or mortality events, of seabirds and marine invertebrates. They gave us fortuitous opportunities to test for resulting natural selection on seasonal migration versus residence.

The pivotal next step was to develop analyses to estimate survival of migrants versus residents throughout the annual cycle, and pinpoint the impacts of extreme climatic events, from the resighting data. Of course, despite our best field efforts, not all individual shags that are actually present in a given area at a given occasion will be seen. The probability that individuals will be missed varied among occasions and locations, to the degree that some possible destinations of migrants were rarely or never surveyed at all. We therefore built statistical models to account for spatio-temporal heterogeneity in missing data, and facilitate unbiased estimation of key parameters. Overall, our models represent spatio-temporal variations in movements and survival of shags, and in our resighting efficiency, across all parts of their range, all seasons and both sexes through the nine-year study. We formulated our model in the Bayesian framework, using the highly efficient open-source software Stan. We took care to code everything in general terms, so that our model can be easily applied to other datasets from different species in future.

When the computer displayed the survival results for the first time, we were stunned by the striking effects. Clear decreases in survival coincided with the extreme winter storms, and with our memories of how harsh the weather had been. The huge differences in adult survival probabilities between residents and migrants were also obvious and fascinating: survival probability was about 0.3 higher in migrants than residents at the seasonal scale, and 0.2 at the annual scale! This represents natural selection (against residence) of an intensity that has rarely been documented for any characteristic. Overall survival impacts were greater on females than males, but the magnitude of selection did not differ between the sexes. These results matched our field intuition. It would appear that migrant destinations provided shelter against extreme easterly storms hitting the east coast of Scotland, while the residency area on and around the Isle of May provided little protection, which led to natural selection through survival.

Our striking results highlight how extreme climatic events could potentially ‘rescue’ a population from the challenges they impose, by causing strong selection for seasonal migration. Now, our ambitions are to quantify inheritance of seasonal migration and examine the full ecological and evolutionary responses to natural selection.

Read the paper

Read the full paper here: Acker, P., Daunt, F., Wanless, S., Burthe, S.J., Newell, M.A., Harris, M.P., Grist, H., Sturgeon, J., Swann, R.L., Gunn, C., Payo‐Payo, A. and Reid, J.M. (2021), Strong survival selection on seasonal migration versus residence induced by extreme climatic events. J Anim Ecol. Accepted Author Manuscript. https://doi.org/10.1111/1365-2656.13410