Blog by Dr Graeme Buchanan, Principal Conservation Scientist, RSPB Centre for Conservation Science

As a conservationist working with maps, I sometimes feel I can relate to those early cartographers who, with limited knowledge of the nature of areas they were depicting, would adorn their maps with phases such as hic sunt dracones, “here be dragons” (or the more common form hic sunt leones, here be lions). Move forward some 500 years, and we still find that in many cases, there is considerable uncertainly over the exact distribution of many species. This lack of knowledge is a real problem for conservationists: how can we protect species and habitat if we don’t know where they are?

In the UK, we are very lucky as our knowledge of the distribution of species of birds is very good. Many years of volunteer survey effort means we have built up an excellent knowledge of the distribution of species during the breeding and non-breeding season. The recent BTO/BirdWatch Ireland/Scottish Ornithologist Club UK atlas covers 2007 to 2011 and gives an outstanding picture of not just where birds occurred in the UK, but, by comparing it to earlier atlases, we can see how distributions changed over time. For example, the atlas clearly shows the northwards expansion of nuthatch and little egret that are generally attributed to climate change. The range contractions of other species, such as lapwing and turtle dove, highlight the need to better understand the ecology of these species if we are to restore their populations.

 

Photo: Change in turtle dove distribution in UK 1968-72 to 2007-2011.

But this level of knowledge is unusual. Across large parts of the planet, we don't know where species do and do not occur. This might not be a surprise for invertebrates, for which many species remain undescribed, but it applies to better known groups of species such as amphibians, birds and mammals too.

This lack of knowledge is a major problem for us conservationists. We need to know the potential distribution of species to assess if they are at risk of extinction. Species with smaller or more fragmented ranges are generally at a greater risk of extinction than species with extensive, continuous distributions. Conservationists also need to know how species' ranges are changing over time to identify those under greatest pressure from human activities. And of course, conservationists also need to know where species occur to target conservation activities. 

Better maps for better conservation

Our knowledge of species' distributions has improved over recent decades. Scientists have compiled records of species to identify the broad bounds of their distributions. These extent of occurrence maps broadly illustrate the area within which a species is known to occur. By drawing around all known occurrences of the species, scientists can estimate distribution. The inclusive nature of this approach means that the maps have low omission errors, which means none of the records of the species are outside the bounds. However, they encompass areas which will not be occupied by the species (commission errors) due to, for example, unsuitable habitat. For example, if you look at the map of the distribution of the common ostrich and zoom in to east Africa, you’ll see that the map includes Nairobi, a vast urban area which is unlikely to be occupied by the species. It also includes Kilimanjaro, a mountain that is too high for the species. 

There are multiple ways in which we can improve our knowledge of species distributions. Conservationists often use species' distribution models. These are mathematical models that compare environmental conditions at locations where species are known to occur (through field work or volunteer surveys) with locations where the species has not been recorded. By describing conditions where the species occurs, we can identify similar areas, and estimate species distributions. To summarise environmental conditions, we have used climate variables such as rainfall and temperature. We have also used land cover and vegetation condition, often derived from satellite remote sensing, for terrestrial species and ocean surface characteristics for marine species.

For many years RSPB scientists have been using species distribution modelling to estimate the distribution of species, and RSPB work has resulted in the identification of new populations of rare birds (liben lark), highlighted areas of importance for rare mammals (pygmy hippos) and helped identify conservation measures for seabirds. However, while these models are exceptionally useful tools for conservation, they are time-consuming to produce. Differences in their accuracy can mean that they are difficult to combine or compare across multiple species. A standardised approach that can be repeated across multiple species across the entire planet is needed.

 

Photo: pygmy hippo recorded on camera trap in the Gola Rainforest National Park in Sierra Leone.

New research on utility of Area of Habitat for the IUCN Red List

A recent paper in “Trends in Ecology and Evolution” describes a method to derive  standardised high-resolution maps that describe where a species might occur, termed Area of Habitat (AOH) maps.  The study, led by Dr Thomas Brooks, Chief Scientist at the International Union for Conservation of Nature (IUCN) brought together scientists from institutions from around the world, including RSPB.

The method takes advantage of technological developments that have occurred in the last decade. There has been an explosion in volume of data about the planet, especially data collected by satellite remote sensing. As a consequence, global maps that describe land cover at a 300m resolution are now available, and data that describe topography at even higher spatial resolutions are free to access.

The study describes the process by which data on land cover and topography are combined with information about species’ habitat, elevation, range, and broad distribution (all needed for a species to be assessed on the IUCN Red List of Threatened Species) to produce Area of Habitat maps.

These maps start with the species’ broad distributions, then exclude areas of unsuitable altitude, and unsuitable land cover. This means that we can be more certain that the species really does occur in the mapped Area of Habitat, reducing commission errors. While there is a risk of increasing omission errors by accidentally excluding areas where the species does occur, the overall balance between commission and omission is expected to be much better. Typically, the Area of Habitat cover about a third to half the area of the original broad distributions.

RSPB has had a long involvement in the development of this method. Ten years ago, together with BirdLife International, Università di Roma La Sapienza in Italy and scientists from the European Commission, we supervised a student undertaking her Masters project using a similar approach to look at the distribution of threatened species in Africa, and how much of their distribution was in covered by protected areas.

This study led on to multiple subsequent collaborative studies, each of which informed conservation prioritisation (and also led to the recruitment of the student to RSPB Centre for Conservation Science). Data quality, processing power, and our knowledge of species themselves have moved on a long way in the intervening decade. The technique has already been applied to generate Area of Habitat maps for more than 20,000 species of mammals, birds, and amphibians, combining data from the IUCN Red List of Threatened Species with remotely-sensed data.

 

Photo: Spatial Measures of Different Aspects of the Distribution of the Azure-Breasted Pitta (Pitta steerii) in the Philippines. From paper in Trends in Ecology and Evolution.

The importance of mapping

Mapping Area of Habitat is not just a scientific exercise: it is also important and timely for conservation policy and practice. In November 2020, the world’s governments will meet in Kunming, China, to negotiate a “post-2020 biodiversity framework” to safeguard and recover life on Earth.

This framework will welcome contributions from all sectors of society towards halting biodiversity loss and advancing restoration. Area of Habitat maps are also being used to identify candidate Key Biodiversity Areas, which are sites of defined importance for conservation of biodiversity. Additionally, Area of Habitat maps are being used in studies looking at scenarios of land use change, allowing a better understanding of the impact of future human activities on species.

Despite some shortcomings, having an agreed method by which we can produce better estimates of species' distributions is therefore a major step forward for global conservation.

 

Brooks, T.M., Pimm, S.L., Akçakaya, H.R., Buchanan, G.M., Butchart, S.H.M., Foden, W., Hilton-Taylor, C., Hoffmann, M., Jenkins, C.J., Joppa, L., Li, B.V., Menon, V., Ocampo-Peñuela N., Rondinini, C. 2019 Measuring Terrestrial Area of Habitat (AOH) and Its Utility for the IUCN Red List. Trends in Ecology and Evolution https://doi.org/10.1016/j.tree.2019.06.009