It started with an invitation and a big challenge. In 2023, my colleague Marcelo M. Rivadeneira asked if I would join him and Guillermo Luna-Jorquera to study how seabirds have responded to climate change through evolutionary time. The group includes albatrosses, petrels, shearwaters, and storm petrels. They are almost exclusively pelagic species and masters of worldwide migration, flying across the entire oceans and seas. However, we still didn’t fully understand how they had coped with past warming and cooling. The literature had clear gaps, and the available methods for the study of historical biogeography felt limited to address such question.
At that exact moment I was developing and applying a novel phylogeographical method (the Geo Model) to primates on land. Here was the perfect opportunity to test it in a completely different system, the open ocean. I jumped at the chance.
Why the Geo Model changes the game.
Common approaches to historical biogeography divide the planet into fixed “boxes” (discrete areas such as the world’s oceans) and assume species simply jump between them. While these methods work well for many questions, they are not well suited to the one we sought to answer. This is because discrete methods do not consider the spherical nature of Earth, ignore continental drift in a realistic way, usually assume the same dispersal ability for ancestral species, and critically, they cannot provide the precise location and environment of ancestral species within the fixed “boxes”.
The Geo Model does something different. It models dispersal as a continuous process across a spherical Earth while explicitly incorporating palaeomaps and variable dispersal rates along phylogenetic branches and over time. For the first time, we could reconstruct ancestral locations as actual distributions of latitude and longitude on ancient oceans, extract the palaeotemperatures those ancestors experienced, and quantify how far and how fast lineages moved through time and across the oceans (Fig. 1).
The Geo Model let us address questions that were previously impossible.
The journey and the rollercoaster of review
The scientific work was exciting, but the review process was tough. We submitted the manuscript and received a rejection after the first round. Reviewers were sceptical about the Geo Model; one even suggested we switch to the commonly used discrete methods because they seemed “more biologically realistic.” Data uncertainty with different phylogenies added another layer of complexity. And there was a broader question from all the reviewers: “This is fascinating history, but what does it mean for today’s climate crisis?”
I was convinced our approach was sound. So, I got ready to do some more work. I ran species distribution models (SDMs) for modern projections, tested multiple phylogenies, strengthened the links between past and present and wrote a detailed appeal. The editor gave us another chance. The revised manuscript came back stronger, and the paper was officially accepted in Nature Climate Change on 23 April 2026, one year after we first submitted it.
What we discovered
Using the Geo Model, we reconstructed the ancestral locations of seabirds across millions of years. We found that these birds did not respond to historical climate change the way many marine ectotherms do (by changing body size). Instead, they changed their geographic range sizes.
- Under rapid local climate change, ranges shrank significantly.
- Smaller ranges were associated with longer dispersal distances.
- These historical patterns align strikingly with what we project for the future. Under severe warming scenarios by 2100, more than 70% of extant seabird species are expected to contract their ranges and disperse even greater distances (Fig. 2).
Shrinking range means that some areas within the species distribution become too warm or too cold, exceeding the climatic conditions the birds can tolerate for survival and reproduction. So, dispersal – by flying away from their original areas - emerge as the key factor to find new habitats that offer optimal conditions.
This evidence shows that conservation efforts must do more than just protect the places where seabirds live today. We also need to safeguard the new areas they may need to move to in the future. The farther away those future suitable habitats are, the harder it will be for seabirds to reach them, especially for those with low dispersal ability.
One analytical “aha!” moment stands out. When I included both temperature and range size as predictors of body mass, the effect of temperature disappeared, and range size emerged as the key driver. That single result opened the door to linking climate → range contraction → dispersal → adaptation in a coherent framework; all made possible because the Geo Model gave us the raw material for hypothesis testing.
Connecting past and future and why it matters for biodiversity
This work builds directly on my earlier PNAS paper on primates, where the same Geo Model overturned the long-held idea that early primates evolved in warm tropical forests. Instead, they radiated through colder and more variable climates. Land or ocean, the message is the same; by modelling continuous dispersal on a dynamic Earth, we can test biogeographic and evolutionary hypotheses with a new level of realism.
My longer-term vision is to make the Geo Model truly accessible to the wider scientific community and new generation of biogeographer and evolutionary biologists. By lowering the barriers to its use, I hope to enable many more researchers to address questions on a dynamic, spherical Earth, ultimately helping us better understand and protect biodiversity in our rapidly changing world.
I am profoundly convinced that methodological innovation like this is essential if we are to understand and protect biodiversity in an era of rapid change. I love this work because every new analysis brings us closer to seeing the living world as it truly evolves: continuously, across a changing planet.
I am deeply grateful to my co-authors Marcelo M. Rivadeneira (species distribution modelling and literature), Chris Venditti (analytical insights and presentation), and Guillermo Luna-Jorquera (natural history expertise and literature). Their trust and collaboration made this possible.
Poster image taken by John J. Harrison. Used under Creative Commons Attribution-Share Alike 3.0 Unported License (CC BY-SA 3.0).