A fire rages through an area of the Amazon

The fight to save the world's ecosystems

Big data - the crunching of huge sets of information - is used in everything from weather forecasting to economic predictions. 

Now, a new dataset hosted by the Museum is helping researchers around the world find and predict patterns of species loss, and reveal what we can do to stop it.

The PREDICTS (Projecting Responses of Ecological Diversity In Changing Terrestrial Systems) project has so far gathered more than 3.2 million biodiversity records. Its goal is to answer one question: how will species diversity around the world change in response to changing landscapes?

Here's how the PREDICTS project is trying to answer this question, how far it has got towards this goal, and why it's such an important question to answer.

A jaguar and an orangutan

A jaguar (Panthera onca) and an orangutan (Pongo pygmaeus) - two animals threatened by habitat loss due to deforestation © Mikadun and jeep2499/Shutterstock.com

Why is the PREDICTS project important?

The world is changing, with more land than ever being converted into cropland, pasture, plantation and urban space. Although this has been happening for centuries, it is only now becoming clear what impact this has had on global species diversity, and what this means for ecosystem stability.

Biodiversity data, such as the number and distribution of individual organisms and different species, is usually only collected for a relatively small area.

It takes a lot of time and resources to collect even this small amount of data. To make global conclusions about the current state and possible future of species diversity, these local datasets need to be combined and analysed together.

The PREDICTS database brings together, for the first time, the largest set of records about the affect of land use how and related pressures on species. It contains more than 3.2 million records, sampled from over 26,000 locations and representing more than 47,000 species. It includes contributions from scientists around the world and is open access, allowing anyone to analyse the data.


Why do we need to conserve biodiversity?

Biodiversity means the variety of plant and animal life in an area of any size, from a square metre of soil to the whole world. High biodiversity - a large range of species - is generally a good thing because it helps to keep ecosystems working effectively. This means that the ecosystem runs itself without human intervention. Also, high biodiversity often makes it more likely that an ecosystem will be able to provide valuable natural resources and other benefits to humans, known as ecosystem services.

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. The higher the variety of species, the more likely the ecosystem is still able to provide the service year in, year out.

These conditions include events such as wildfires and droughts, as well as more permanent changes, for example climate change creating increasingly hotter summers.

Pollination is one example of a vital ecosystem function that could be under threat as a result of shrinking biodiversity. With 35% of global food production in some way dependent on animal pollination, a loss of pollinators could mean lower crop yields and less food worldwide.

Bee pollinating a white flower

Deforestation for agriculture and urbanisation puts pressure on vital ecosystems, such as those that provide crops for food and clothing © Yuri Kravchenko/Shutterstock.com

What exactly does the data tell us?

The data can be used in a number of ways, depending on what each analysis is trying to determine.

For example, the data has been used to estimate the average loss of species from local ecosystems worldwide. The research team estimates that average species numbers have fallen by 13.6% since 1500 because of natural habitats being turned into fields, pastures and towns.

This means that if you were to go out and sample a site anywhere in the world, on average you would find 13.6% fewer species than you would have done in the year 1500. As this in an average, ecosystems in some parts of the world have experienced much higher losses. For example, sites in western Europe have lost more than 20-30% of their species on average since the Industrial Revolution (late eighteenth and nineteenth centuries).

Research using the database has also found that human land use has driven the level of biodiversity down to a dangerous level across 58% of the world's land surface.

It's not all bad news, however. The database has also been used to determine that protected areas - regions across the world where certain human activities are restricted - are helping to conserve species. On average, there are 11% more species inside protected areas than outside.

Map showing locations of sites included in the PREDICTS project

The PREDICTS database contains more than 3.2 million records, sampled from over 26,000 locations and representing more than 47,000 species

Where is most biodiversity being lost?

An analysis by the PREDICTS team found that grasslands, savannahs (mixed grass and woodlands) and scrublands have been most affected by biodiversity loss, followed closely by many of the world's forests and woodlands.

Worryingly, biodiversity hotspots - areas that are the only home for large numbers of species, and with high levels of habitat loss - are among the regions badly affected. Of the 34 hotspots identified worldwide, 22 suffer from unsafe levels of biodiversity loss.

'Unsafe' is defined as an average species loss of more than 20 percent. This level is widely considered to be a likely tipping point at which ecosystem services are compromised. An ecosystem is a bit like a tower of Jenga: it can usually survive the loss of a few pieces, but removing further pieces becomes riskier and riskier and in the end it collapses.

It's estimated that over a quarter of Earth's land surface has already exceeded the 20% 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.

A charred tree stump in a deforested area of the Amazon

A charred tree stump in an area of the Amazon cleared for cattle ranching © Frontpage/Shutterstock.com

How can the data help inform policy and decisions about conservation?

The PREDICTS data has been used to suggest how biodiversity has changed in response to the destruction of natural habitats so far. By analysing these patterns, the team have also been able to predict how it might respond to future changes.

The researchers looked at four different scenarios for the future of climate change, ranging from 'business as usual' to different approaches to mitigation, where we make a coordinated effort to reduce the impact of climate change. They then predicted how global biodiversity would respond to each of these scenarios.

Forecasting global species diversity change

A graph showing the net change in global land-based species diversity over time, and future scenarios. The red line represents the continued loss of terrestrial species diversity that would be likely if forests are continually exploited. The blue line shows the potential increase in biodiversity if forested areas are protected.

What the team found was that the last 50 years of damage to biodiversity could be undone if we mitigate climate change by using carbon markets to fully value high-biodiversity forest areas.

This means that forests would be protected because they have a monetary value based on their ability to absorb carbon dioxide and prevent further climate change. At the moment, because forests are not valued properly, it is often profitable to convert them to agriculture.

Under that best-case scenario, 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, species diversity would reduce by a further 3.4% by 2100.

PREDICTS leader Prof Andy Purvis says, 'Decision makers worry a lot about economic recessions, but an ecological recession could have even worse consequences.'

The world is changing, but only a truly global dataset can tell us how much. The PREDICTS database does just that for the world's biodiversity - the variety of species that support the function of ecosystems and ultimately provide us with services like clean water, sustainable resources and crop pollination.

The PREDICTS team have already estimated that on average 13.6% of the world's species have been lost since 1500. But they also highlight the best ways to prevent further losses - by maintaining and expanding conservation areas and by valuing high-biodiversity areas.


More about PREDICTS


Science news

Get the latest news about the Museum's scientific research and unique collections.