Aurélie C. Shapiro
Remote Sensing Specialist
World Wildlife Fund US
The main drivers of environmental change and loss of biodiversity can be attributed to human impacts – habitat clearing and degradation, pollution, fragmentation, hunting, exploitation and land conversion, to name a few (Mace et al., 1998).
The effect of this human footprint is drastic, and highly correlated with proximity to human settlements and access, notably via roads and other infrastructure which allow humans to interact with natural areas and degrade them (Forman, 2000; Riiters and Wickham, 2003; Kareiva et al., 2007; Sanderson et al., 2002). GIS is the perfect tool to map and assess the potential human impacts on ecosystems worldwide in order to generate proxies for threat to landscapes and the relative cost for conservation. The World Wildlife Fund (WWF) works closely with communities who are dependent on fragile, threatened natural resources, therefore identifying the potential overlap of human activities and the natural environment is a crucial step in that effort, essentially, the path to a sustainable earth.
Mapping human access or anthropogenic threat is an important assessment to determine future targets for conservation investments in terms of their vulnerability and persistence, as well as initial cost for conservation (e.g. restoration: a degraded environment is more costly to keep intact or restore). In addition, human access and influence is important to monitor over time to flag new development or settlement expansion that could threaten ecosystems, increase the spread of disease as well as decrease ecosystem services (O' Sullivan et al., 2008; Foley et al., 2007; Wolfe et al., 2005; Sullivan et al., 2008). The World Wildlife Fund's Coastal East Africa priority place (Figure 1) extends over 4,600 km from Southern Somalia to South Africa. This diverse region includes coastal forests, the Eastern Arc Mountains and Miombo woodlands, and habitats for the largest number of globally important threatened and endemic species in Africa, while also being home to some of the poorest human communities in the world. People in this region live on the edge of survival, facing daily hardships, wars and natural disasters, all while relying mainly on natural resources for food, health, shelter and livelihood. WWF has projects throughout the region, including the Ruvuma Wilderness Area, where elephants and other species migrate throughout the seasons following sources of water, resulting in human-elephant conflict. WWF is assessing corridor areas to mitigate this conflict. Further south, the Primeiras and Segundas Archipelago Reserve has been proposed to protect community fishing zones and enhance the local economies.
A snapshot of potential human threat and access was calculated for the coastal east African landscape, similar to the human footprint, but more specific to available local and regional data sets. This analysis provides not only valuable information on the 'remoteness' or wilderness of areas, but also impending threats and the ability to update data with planned or proposed development projects in order to
"Mapping human access or anthropogenic threat is important to determine future targets for conservation investments"
manage and mitigate new pressures on ecosystems. This information will help manage vulnerable natural resources that play vital role for local communities. The analysis of potential threat and access was based on a segment of The Nature Conservancy's Threats Assessment version 1.1 designed for South America (Jarvis et al., 2006). Point locations of towns and cities were obtained to determine the general presence of humans in the landscape, or main sources of threat. A transportation network was defined using available road data, railroads and navigable rivers. Major cities, roads, and rivers are shown in Figure 2, along with protected areas.
A friction, or velocity layer was derived to determine human movement throughout the landscape on either the transportation network or through different land cover types, using the Global Land Cover 2000 (JRC, 2000), which was the best consistent data layer for the entire Coastal East Africa region. Velocity for human travel was estimated within a 1km grid throughout the whole region, on both the transportation network and on land, by estimating the speed in which it would take to travel across that cell. Elevation and slope from hydrosheds data (Lehner et al., 2008) were used to 'slow' velocities to simulate the difficulty in travelling through complex terrain, mountain obstacles or higher altitudes.
To model access, or human movement throughout the landscape, the velocity was converted to friction, which is the time it takes a person to travel across each cell. This is based onthe land cover (dense forest is more difficult to cross than open fields), the type of road (paved or a path), or whether the cell is a railroad or navigable river. The friction and town locations were inputs for the cost distance function in ArcGIS, which estimates the effective cost it would require to
"Population density plays an important role in the magnitude of threat in a region"
travel from any source (city, town) to any point in the landscape. The inverse of this layer is the human threat: it defines areas that are more likely remote and intact (higher cost to access), as well as those that are more accessible and more degraded (low cost).
Since population density also plays a role in the magnitude of threat in a region, a human density surface was created using the best available population data (CIESIN settlement points) using kernel density. The two components, population density and the cost layer were rescaled to 0-1 and added together for a total combined human threat (Figure 3). The value is unitless, with a relative value between 0 and 2. This type of data is commonly used to design corridors, protected areas, community reserves and for other conservation land planning in the region, where remote, presumably more intact areas would be of lower cost for conservation and more feasible to adequately protect.
It has often been the case that protected areas are located in regions with relatively little access – there is less potential conflict and land is easier to acquire. Whether this remains true however, depends on whether parks are adequately preventing new infrastructure within strictly protected areas or buffer zones.
In coastal east Africa, a combination of these factors is the case. According to the map, most of the parks tend to be located in large areas of low threat (green), while some newer protected areas like the terrestrial buffer zone around the Primeiras and Segundas Archipelago Reserve already have existing infrastructure and populated settlements.
This is precisely why updating this analysis with new land cover, planned or implemented infrastructure projects will identify the path of human threat over time. In addition, WWF and partners can advocate for better environmental planning for new projects that steer clear of endangered species habitat, corridors or other important ecosystems.
This map is simply the first step in long term monitoring and conservation planning in the region, using a standard methodology that can be reproduced elsewhere.