Big data on biodiversity

A tea estate in Agrapatana, Sri Lanka. Removal of primary vegetation for agriculture has a huge impact on local biodiversity. 

From ecologists in Ecuador to botanists in Bognor Regis, scientists around the world are collecting valuable data on species under pressure from human activity.

A major collaborative project brings those records together into a huge database of evidence - the largest survey ever completed on the impact of humanity on local biodiversity.

Project lead and Museum scientist Prof Andy Purvis says:

‘We know that the landscape is going to change a lot in the future as the population grows, but we haven’t really known how biodiversity will change in response. With PREDICTS we’re building global models that help us to predict how land-use change will affect local biodiversity - and us - in the future.’

The Projecting Responses of Ecological Diversity In Changing Terrestrial Systems (PREDICTS) team is tracing the impact of human activities like land-use change, pollution and infrastructure on local ecosystems.

More worms, please

A vast terrestrial assemblage database of unprecedented geographic and taxonomic coverage, the database currently contains records from 90 countries and 450 scientific papers, representing more than 40,000 species - that’s around 1.5 per cent of species formally described by science. While that might not seem like a lot, it’s an incredible achievement. And the counter is still ticking upward.

PREDICTS researcher Dr Lawrence Hudson says:

‘The database continues to increase in size. That’s because of the generosity of hundreds of scientists around the world and the hard work put in to get those records into the database.’

Underrepresented groups (like annelid worms) and regions (like boreal forests) are specifically targeted by the team, ensuring that the database becomes more indicative of natural diversity through time.

All of the PREDICTS data, currently more than 2.5 million biodiversity records, will be released onto the Museum’s Data Portal later this year and made accessible to researchers around the world.

Crossing the threshold

The first global analysis of PREDICTS data, published in Nature in April 2015, suggests that the conversion of land for human activity has already caused a significant worldwide loss in local species diversity. Whether for cropland, pasture, plantation forest or urban use, it’s all bad news for biodiversity.

Prof Purvis explains:

‘What the figures show is that if you were to go out and sample a site, anywhere in the world, on average you'd find 13.6 per cent fewer species than you would've done in 1500. And that appears to be because of major land-use changes by humans.’

That figure is a global average, so local biodiversity in some areas is still relatively intact, but others - including Western Europe - have experienced losses in excess of 20-30 per cent since the industrial revolution.

The worldwide pattern of local species richness change since 1500


Unfortunately for humanity, 20 per cent is widely considered the tipping point at which biodiversity’s contribution to ecosystem services is compromised. These ‘services’ include raw materials for building and clothing, food sources, control of pests and diseases, clean air and water, breaking down waste products and recycling materials.

It’s estimated that over a quarter of Earth's land surface has already exceeded the 20 per cent threshold of biodiversity loss. Those ecosystems may only be continuing to function reliably - if they are doing so - because of the extra effort we put in, such as spraying crops against pests or purifying our own water.

Recouping species losses

The future of global biodiversity is still potentially in our hands, according to the team’s analysis. They projected biodiversity’s response under four different climate change mitigation scenarios described by the International Panel on Climate Change (IPCC), which are our best estimates of how the global landscape will change over the next century.

What the team found was surprising, admits Prof Purvis:

‘I had expected these scenarios to be variations on a rather dark theme. But we could actually undo the last 50 years of damage if we mitigate against climate change by using carbon markets so that we fully value high-biodiversity forest areas.’

Under that scenario, known as MINICAM 4.5, forested areas are protected as economically valuable resources, and the great majority of countries see their biodiversity increase over the rest of the century.

At the bleaker end of the spectrum, if humanity does absolutely nothing to counteract climate change, we’d lose a further 3.4 per cent of biodiversity by 2100. This figure does not even take into account the inevitably huge impact of climate change itself, should we fail to mitigate against it.

Under a business-as-usual scenario the vast majority of countries see a decrease in their species richness over the next hundred years. If instead humanity protects forested areas and supports carbon markets (MINICAM 4.5), almost all countries could actually gain back biodiversity by the end of the century.


Delving deeper into the database

With a vast geographical, ecological and taxonomic range, PREDICTS has the potential to fuel a range of future analyses.

A new three-year project will take a slightly different approach to similar questions, by gathering data from sites before and after a specific land-use change. This will reveal direct links between land-use changes and ecosystem responses, as well as how long those responses take and how they vary by organism. It will also let the team test some assumptions they’ve been relying on up to now. 

Other upcoming projects will consider the impact of human pressures on measures of biodiversity that go well beyond numbers of species, and how effectively protected areas such as national parks preserve biodiversity. The team will also continue a well established collaboration with Microsoft Research’s computational ecology lab in Cambridge.

The PREDICTS database continues to grow, becoming more representative of nature through time. It’s impossible to say how valuable it may become in the future as we strive to retain our planet’s biodiversity.

As for biodiversity itself, Dr Hudson sees it as more than just a nebulous scientific concept:

‘As a species, as a civilisation, we're completely dependent upon the natural world. We can't live in isolation from nature. Studying biodiversity is fundamentally important for our survival as a species.’

If you study local biodiversity, you can get involved on the PREDICTS project website.

© Domenico Tozzi

Project team

  • Andy Purvis
  • Tim Newbold
  • Sara Contu
  • Samantha Hill
  • Adriana De Palma
  • Helen Phillips
  • Victoria Burton

External collaborators

Related publications

Newbold T et al (2015) Global land-use impacts on local terrestrial biodiversity. Nature.

Biodiversity initiative

Find out about our research initiative focused on the diversity of tropical forests.


Supported by

Biodiversity research

We are creating molecular and digital tools to explore undiscovered biodiversity

Diversity and informatics research

Researching undiscovered diversity in megadiverse systems using big data


The Museum's 80 million specimens form the world’s most important natural history collection