The macroevolutionary consequences of trait correlations

Lead Research Organisation: University of Sheffield
Department Name: Animal and Plant Sciences


The diversity of Life on Earth is immense. Over millions of years, evolution has generated a staggering variety of species. When we look more closely, however, we begin to notice some surprising patterns. Some groups have far more species than others, for example there are over 350,000 species of beetle but only a few species of elephant. Other groups exhibit diverse shapes and sizes, for example the Hawaiian honeycreepers, whereas others are fairly uniform, for example mice and rats. These patterns suggest that there are constraints, or limitations, on the numbers of species and their variety of form and function, i.e. biological diversity is not evenly distributed across the tree of life. Why this is so remains unresolved and is a fundamental question in evolutionary biology.

Here we focus on a previously unexplored explanation for this uneven distribution of diversity: correlations among species traits. Species traits can be almost any aspect of physical appearance (such as body size or eye colour), behaviour, or life history (size vs. number of offspring). These traits are not always independent of one another, for example investing in more offspring often results in smaller offspring. In such cases we say that the traits involved are correlated. Where traits are correlated, evolution in one trait may promote or prevent evolution in another.

Differences in the strength of trait correlations in different groups of species might determine variation between and among groups of related species. Where correlations among traits are strong these may prevent evolution of very different shapes and sizes. Therefore the diversity of form and function might be low, and the number of closely-related species should be few. Over millions of years these trait correlations might become weaker or even disappear, allowing species to evolve in new ways or multiply in number. These ideas suggest that trait correlations are important in our understanding of the diversity of life.

Our proposal aims to understand how trait correlations can affect the processes that determine patterns of diversity (both in species number and in form and function) across the tree of life. We focus on traits related to food type and acquisition in birds. The size and shape of bird beaks can reflect dietary requirements and foraging methods. When traits are correlated, some combinations may never occur together, for example, it may not be possible to evolve beaks that are simultaneously long and curved like a curlew, and wide and deep like a duck.

To test how trait correlations can affect the processes that determine patterns of diversity across the tree of life we need to measure many individuals from many different species. It would be a near-impossible challenge to do this in the field. Instead we use specimens from museums. Around the world museums house billions of specimens, often from historical collections over 100 years old. The Natural History Museum, London has around 80 million specimens. These collections are a rich source of information on nearly all species of birds (over 10,000 species). We will generate data from these collections, including 3D scans of bird beaks. We will then use these data, along with state-of-the-art analysis methods, to evaluate trait correlations among and within species, and among higher taxonomic groupings such as families and orders.

By using museum specimens and a novel view of the importance of trait correlations, our research will provide new insights into how and why life diversifies. Not only will this research address a fundamental debate in evolutionary biology, but it will also create valuable bird datasets for future researchers, and highlight the importance of natural history collections for cutting edge research.

Planned Impact

Who might benefit from this research?
The research in this proposal will have two main impacts. It will (1) promote natural history museum collections and museum-based research; and (2) enhance student training in quantitative skills. Impact 1 will benefit (i) museums; (ii) the people of the UK; (iii) the general public; and (iv) science policy makers. Impact 2 will benefit (v) students and postdoctoral researchers; (vi) employers; and (vii) the UK economy.

How might they benefit from this research?
(i) Museums. Natural history museum collections are a hugely undervalued resource that are often unappreciated by the general public, funding bodies, policy makers and governments. If we wish to preserve the valuable collections held in our UK museums, we need to promote their utility much more widely. We will demonstrate the value of natural history museum collections and highlight how specimens collected a hundred years ago are still relevant to cutting-edge research today. This will benefit museums by providing evidence of their utility and impact when funding is in question. (ii) People of the UK. Many of the specimens held in UK museums were collected hundreds of years ago, by historical figures like Darwin and Captain Cook. By demonstrating the value of natural history museum collections we hope to preserve them for future generations, thus benefitting the people of the UK by retaining their cultural heritage. (iii) General public. Public trust in science/scientists is at an all time low. Yet people remain fascinated by the natural world. We will engage in an extensive series of outreach events, media coverage and online activities including public talks, events, exhibitions, press releases, blogging and social media directed at interacting with the public, both in the UK and worldwide. This will benefit the general public allowing them to interact directly with scientists and increase their knowledge of the natural world and scientific methods. (iv) Policy makers. The aforementioned public engagement program will increase public understanding of science, paving the way for smoother interactions between policy makers and the general public. (v) Students. Academic jobs are becoming increasingly scarce, and PhD students and postdoctoral researchers are looking to diversify their skills to find jobs outside of academia. Both the public and private sector are actively seeking to employ people with strong quantitative and data science skills. We will train students and postdoctoral researchers in these skills via workshops and online teaching materials which will benefit them by increasing their employability, and help with their current projects. (iv) Employers. The training in quantitative skills that we provide to students and postdoctoral students will benefit employers by providing a pool of applicants with programming, database management skills, and general computing skills. (vii) UK economy. There will be longer term economic benefits to the UK when the students we train transfer their data science skills to the workplace, ensuring the UK remains competitive in this area.

All three participants have extensive experience in delivering impact of this kind. NC and GHT are heavily involved with public outreach at NHM, NHM Tring and the Alfred Denny Museum. GHT has extensive experience with writing press releases and his 2012 Nature paper on bird phylogeny attracted a great deal of media attention. Between them, the participants have supervised ten PhD students, five postdocs and over 20 Masters students in the last five years. APB and NC are instructors at R4All and all three regularly teach R, statistics and/or phylogenetic comparative methods. NC also teaches a BBSRC STARS course on computational skills and recently edited a Guide to Reproducible Code for the British Ecological Society (


10 25 50