ZSL’s journal Remote Sensing in Ecology and Conservation publishes new research at the interface of remote sensing science and conservation. A recent paper in the journal explains how remote sensing is being used to identify IBAs, sites of international significance for the conservation of birds (and other taxa), that are at greatest risk of deforestation.
For years, conservationists have been calling out for an easy-to-use method of monitoring land-cover change across sites of conservation importance. In 2013 Hansen et al. (Science 342, 850–853) produced a 30m resolution global map of tree cover, and tree cover change. Despite acknowledged limitations of the map, it provided an opportunity for conservationists to assess and visualise broad patterns of tree loss at a global scale.
Analysis of these data on sites of conservation concern would enable conservationists to identify conservation priorities to which effort could be targeted. Examination of correlates of change could enable conservationists to identify the characteristics of sites that make them susceptible to forest loss, enabling pro active conservation. But how to do this for such a large data set? The solution turned out to be through a collaboration between a physicist eager to help conservation, the secretariat of BirdLife International, the global partnership of conservation NGOs, and the Royal Society for the Protection of Birds, the UK BirdLife partner. Oh, and a massive amount of help from Google Earth Engine’s staff and their cloud processing power...
The Cloud Forest
Google Earth Engine is a cloud-computing environment dedicated to geospatial analysis that allows rapid processing of vast amounts of data. Best of all, it does not require the developer to be an expert on parallel programming paradigms. Basic knowledge of Python or Java Script language is enough to perform otherwise complicated analyses. Suffice to say that all our output was generated with little more than 40 lines of code!
From the petabytes of available on Google Earth Engine data sets, we focused on Global Forest Change map provided by Hansen et al. It allowed the assessment of tree-cover change between 2000 and 2012 within each of the 7,700 Important Bird and Biodiversity Areas (IBAs) which hold forest birds. These sites are a subset of the global network of over 12,000 IBAs.
Action and Reaction
After the extraction of these data, we identified correlates of forest loss. This enabled us to identify which characteristics of IBAs are associated with them being at greatest threat from deforestation. This knowledge can help target resources to sites with attributes that point to them being at greatest risk. Think active intervention. But the most urgent use for these data is in identifying IBAs being damaged and degraded now, in order to call in the cavalry before, fingers crossed, it is too late – reactive conservation. Conservation resources are limited, so it is important that the sites to which resources are targeted first are those in most immediate need of support.
Our data have already fed directly into BirdLife’s vital “IBAs in Danger” initiative (http://www.birdlife.org/datazone/info/IBAsInDanger), which aims to highlight those IBAs around the world in direst need of conservation intervention if they, and the species that they support, are to survive. These represent the most threatened of the most important sites for birds and other biodiversity across the globe.
Point of use
These data have now been circulated to regional BirdLife secretariats in Southern America, Oceania, Europe, Asia and Africa where they are being used close to the point of use. As detailed in our paper, feedback from this process has highlighted some issues with these data. For example, it is widely recognised that the Hansen et al data do not differentiate between natural and plantation tree cover. It transpired that the site which experienced the highest proportional loss of trees in the study had recently had a plantation removed. Step the cavalry down. Such local knowledge is essential if these data are to be exploited to their full potential.
In Australia the loss of forest confirmed concerns regarding the trends in natural habitat loss as a result of increasing occurrences of fire in two sites already identified as IBAs in Danger. These data highlighted another six IBAs across Australia where over 25% of tree coverage had declined during the survey period. BirdLife Australia staff are checking to see what impact this is likely to have on species at these sites.
But caution is needed when using these data alone - field surveys are still needed. We found that the forest cover map for many parts of Fiji, including the third largest island Taveuni, appears unable to differentiate between forest and cultivated habitats. Consequently, the change in the extent of forest cover that we reported is unlikely to be an accurate assessment of what has happened to IBAs here. Finally there is no information on forest cover for most of the Polynesian and Micronesian Islands. While the area of forests in these countries is small the proportion of globally threatened species that they contain is disproportionately high. We have recently provided feedback to the data producers on these issues.
Despite these issues, the fact that these data have already been used to inform conservation priorities illustrates not only how important they are, but also how remote-sensing data can be used to serve the needs of the conservation community. At last, the recognised potential of remote sensing to global conservation monitoring and priority setting is being realised.
By Graeme Buchanan, Lincoln Fishpool, Lukasz Tracewski and Mark O’Brien
Graeme Buchanan, Lincoln Fishpool and Łucasz Tracewski were authors on the paper. Mark O’Brien is the Regional Programme Co-ordinator, for BirdLife Pacific, and here provides a user perspective on the analysis.
Select a blog
Our people are our greatest asset and we realise our vision for a world where wildlife thrives through their ideas, skills and passion. An inspired, informed and empowered community of people work, study and volunteer together at ZSL.
At ZSL, a key area of our work is the employment of Nature-based Solutions – an approach which both adapt to and mitigates the impacts of climate change. These Solutions, which include habitat protection and restoration, are low-cost yet high-impact, and provide multiple benefits to people and wildlife. We ensure that biodiversity recovery is at the heart of nature-based solutions.
A blog for lovers of ZSL London Zoo, bringing you extraordinary animal facts and exclusive access to the world's oldest scientific zoo.
Do you love wildlife? Discover more about our amazing animals at the UK's biggest zoo!
We're working around the world to conserve animals and their habitats, find out more about our latest achievements.
From the field to the lab, catch up with the scientists on the cutting edge of conservation biology at ZSL’s Institute of Zoology.
A day in Discovery and Learning at ZSL is never dull! The team tell us all about the exciting sessions for school children, as well as work further afield.
Every month, one of the pieces held in ZSL’s Library and at ZSL Whipsnade Zoo will feature here as Artefact of the Month.
Read testimonials from our Members and extracts from ZSL's award winning members' magazine, Wild About.
The Chagos archipelago is a rare haven for marine biodiversity. Hear from the team about our projects to protect the environments in the British Indian Ocean Territory (BIOT).
ZSL works across Asia, from the famous national parks of Nepal to marine protected areas in the Philippines. Read the latest updates on our conservation.
An Open Access journal for research at the interface of remote sensing, ecology and conservation.