The Scottish isles of St Kilda, off the west coast of the Outer Hebrides, have an important place in my heart. It was on St Kilda where I first realised that not all sheep are boring, where I sustained my first fieldwork-related injury (a broken hip caused by an impact during a sheep-chasing incident!), where I successfully ran my first Research Council grant, and where I met my wife. It was also on St Kilda where I gained my first taste of a long-term ecological study, and where I realised what a tremendous effort is required to ensure their sustained persistence. On my first visit (lambing 1993), the St Kilda Soay sheep project was still in its infancy, having been conceived in its current form by Tim Clutton-Brock and Steve Albon in 1985. This year, the project celebrates its 30th anniversary and to mark this milestone the team organised a programme of public talks at the Royal Society of Edinburgh.
I am sure that a summary of all the excellent presentations at the meeting will appear elsewhere, so here rather than repeat this, I wanted to reflect on some of the broader issues that these talks, and the project as a whole, raises about what makes a successful long-term study (with a special nod to Dave Coltman for the inspiration).
But first, is this really a long-term study and has it really been that successful?
Well, 30 years certainly feels like a long time, but the Soay sheep project is a spring chicken in comparison to some other long-term animal ecology studies – think great tits at Wytham Wood, chimpanzees at Gombe, elephants at Amboseli, Darwin’s finches on Daphne Major, aphids at Rothamsted, red deer on Rum, guillemots on Skomer, to name but a few [see Begging for funding]. However, if success if gauged in terms of research publications then the project compares favourably with the best of these – over the last three decades more than 150 papers have been published from the project, accumulating between them more than 5,000 citations (h-index = 42).
Another useful metric of success is the number of young scientists the project has trained and mentored. This is more difficult to quantify, but at a best guess the project has probably trained several dozen doctoral students and post-doctoral scientists (including two current, and one previous, JAE senior editors!), not to mention scores of undergraduates and Masters students who have benefited from Soay sheep data and samples, and the literally hundreds of volunteers who have gained a valuable life experience. Of course, the Soay sheep project is not unique in either of these regards – indeed the production of lots of high-quality publications and young scientists is a well characterised output from most long-term studies, which is why they should be promoted and protected, as discussed previously on this blog.
So, why has the Soay sheep project been so successful?
Generous leadership: Over the years, a hallmark of the Soay sheep project has been that it has welcomed in new collaborators who bring in fresh skills and perspectives. What initially started out as an analysis of the causes and consequences of the unstable sheep population dynamics, quickly incorporated new questions about sheep genetics and the inheritance of traits, the role of parasites, vegetation dynamics, sheep behaviour, demography, quantitative genetics, genomics, immunology, ageing and physiology. It would have been easy for the project leaders to guard access to such a valuable and unique ecological resource, but by inviting new scientists to the party and encouraging ex-students and post-docs to develop their own areas for development, the Soay sheep project has continued to be at the vanguard of ecological science, with new collaborators adding value to ongoing studies rather than competing with them. That is not to say that it has always been smooth sailing (either metaphorically or literally – access to the islands requires a sea journey of at least 40 miles on often choppy seas!), or that personalities and egos have not clashed at times over the years. But under the considered and generous leadership, of first Tim Clutton-Brock and then Josephine Pemberton, the long-term future and development of the project has always taken centre stage.
The long game: Another reason for the project’s continued success is that whilst grant funding is typically short-term (usually three years), the outlook of the project has always been longer-term, with strategic planning of grant applications ensuring continual Research Council funding for the entire 30 years of the project – a quite spectacular feat! This has been made possible only by the combined efforts of lots of individuals and not just the nominal leaders of the project. Science funding is typically fickle, with research ideas coming in and out of fashion. An important bi-product of generous leadership is that there can be a pluralist approach to funding, with multiple applications for core funding being possible due to the diverse nature of the study and the questions that are currently considered ‘sexy’. Indeed, over the years the ‘core’ long-term monitoring of the Soay sheep project has been funded by grants to at least half-a-dozen different individuals.
The project has also taken a long-term approach to its understanding of the key questions being tackled. As ecologists, we are well aware that we can never fully understand ecological systems, and that all we can ever really do is to establish our best guess at the ‘truth’, which we must then update in light of new information. This is perhaps best illustrated by considering how our understanding of Soay sheep population dynamics has evolved since the project began. In 1992, just before I first visited St Kilda, Bryan Grenfell and others used the available data (6 years of high-quality census data from Village Bay) to argue that the unstable dynamics of the sheep were in fact intrinsic cycles driven by overcompensating density-dependent mortality. After I joined the project, we updated that assessment (based on 40 years of whole island census data from Hirta and neighbouring Boreray) to argue that the dynamics were not in fact cyclical but were due to pronounced threshold effects with population crashes occurring in years above some critical sheep density, the depth of which depended on winter weather (as illustrated by synchronous crashes occurring on isolated but neighbouring islands – the Moran effect). A few years later, Tim Coulson and colleagues showed that population age-structure was also important. Oh, and by the way, the long-term trend is for the sheep population size to increase and for the sheep themselves to get smaller, probably due to climate change. And to understand the mechanisms underpinning all this, we also need to consider the interaction between the sheep and their food supply: after more than 20 years of twice-yearly vegetation monitoring by Mick Crawley and students, we are only now getting close to understanding how the sheep and their food supply interact with each other and climate. The point is this: only by combining long-term data collection with a multi-disciplinary research agenda can we hope to tackle these ecosystem-scale interactions and the impact of large-scale phenomena such as global warming.
Continuity: Finally, another factor contributing to the success of the Soay sheep project is the continuity provided not just by its leadership and long-term collaborators, but also by its field staff and data curators. Jill Pilkington MBE has worked with the sheep project for over 20 years now, visiting the island for weeks at a time every spring, summer and autumn, come rain or shine. Ian Stevenson completed his PhD on St Kilda in 1994 and for the past 15 years has developed and managed the Soay sheep database, latterly in his role as MD of Sunadal Data Solutions . Their continued contribution to the project (and in Ian’s case to many other long-term studies) has not only allowed new methods and approaches to be developed and refined to work like clockwork, building on previous successes and rectifying previous errors, but has also provided an invaluable resource for new students and collaborators to mine. It is also worth noting here the important long-term support provided by the National Trust for Scotland, which owns the St Kilda, and Scottish Natural Heritage, which has oversight of conservation on the islands; without their continued understanding of the value of scientific research, this project could have died a long time ago.
I am not sure if the traits I have highlighted above are common to other successful long-term studies or if the Soay sheep project is unique. But either way, I think it provides a valuable example for other potential long-term studies to follow.
Senior Editor (@spodoptera007)