Why are non-governmental environmental organizations mapping intact forest landscapes?
Unfortunately there is no governmental body carrying out inventories of remaining intact landscapes globally - be it of forests or other land ecosystems. The destruction of these areas is happening at a quick and increasing rate, in particular in the tropics and in some temperate and boreal regions, creating the urgent need for such maps.

Why was the presence of infrastructure rather than its density used to evaluate the fragmentation of natural areas?
In many cases even one road can lead to a fundamental increase of human impact on natural areas such as increased fire frequency, poaching, logging and burning for agriculture, etc. This is especially the case with "pioneer development" of large natural areas, which still takes place in many regions of the world where ancient forests are concentrated. Using the density of the road network as a criterion of fragmentation does not always reflect the reality of such impacts.

Why is the size of buffer zones along all types of infrastructure 1 kilometer?
Allocation of buffer zones along all types of infrastructure (roads, waterways, settlements etc.) of one size (1 kilometer) is a simplification, which is necessary on a global scale. It would be reasonable to use buffer zones of different sizes for different types of infrastructure. However, a clear distinction is not always possible using satellite images and global maps showing all types of infrastructure in a consistent manner don't exist.

It's noteworthy, however, that in this work the size variance of buffer zones (up to several kilometers) is not significant or visible on the map. There are rare exceptional cases where changing the size of buffer zones could influence the identification of an area as intact forest landscape.

Why is the minimal area of the identified intact landscapes 50 thousand hectares, and why does it not change in different regions and vegetation zones?
The conservation of large natural areas including forests is important for a number of reasons. It helps to maintain ecological integrity of ecosystems by maintaining ecosystem composition, structures, and functions across spatial scales through time, provides the large habitat areas needed for the survival of viable large mammal populations (which can often be identified as umbrella species, which means protecting their habitat will likely protect a wide range of co-existing species in the same habitat, which may be lesser known and difficult to protect otherwise); help to minimize dangerous "edge effects" (like encroachment of invasive and alien species, poaching, increased fire and wind disturbance, biogeoclimatic changes, etc.); and is necessary to protect watershed drainage areas.

The 50 thousand hectare threshold ensures the inclusion of most intact areas globally. Identification of smaller natural areas can be important too, but would be subject of another major research study.

Why is there a time limit for detected disturbances?
There are several reasons for this. For thousands of years, traditional human activities have influenced the global environment, including forests. In many cases it is impossible to separate the long-term natural and human impacts on even the most remote and isolated forests from each other, especially where the indigenous peoples have traditionally lived in forests and used forest resources.

The goal of this work is to detect disturbances created by high-intensity forms of use, which usually differ from traditional forms of natural resource use over thousands of years. That is why a timeframe coinciding with the formation of modern industrial economies was used.

Also it is important to note that most satellite images cannot be used to detect one-off disturbances (for example, a disturbance which took place more than 60-70 years ago in a boreal forest or 30-40 years ago in tropical and temperate forest, even considering the supplementary use of 10-year old Landsat TM images).

Why were the burnt sites adjacent to infrastructure regarded as disturbed areas?
With remote sensing techniques it is mostly impossible to tell whether the cause of a forest fire was human-induced or natural. In areas with new infrastructure fire frequency increases because of the increased number of people in the area, and thus the potential causes of fire grow.

Moreover, the succession of forest vegetation in fire scars adjacent to infrastructure may differ from the natural succession in fire scars which are completely surrounded by intact areas as the possibility of new fires, encroachment of invasive and/or alien species or the use of burnt areas for agricultural purposes (especially in the tropics) increases. Thus, burnt areas adjacent to infrastructure were excluded from intact forest landscapes as it was assumed that they were 1) probably human-caused and 2) likely created an altered successional pathway which likely differs from the natural forest succession of remote fire scars embedded in a natural surrounding.

Does the map show all the disturbances and threats to ancient forests?
No, it does not. Some threats cannot be revealed by remote sensing methods and/or because some areas impacted have diffuse boundaries. An example of a threat that is not shown in the map is intensive unregulated hunting, which leads to the decline of large mammal populations in the tropical forest regions of Central Africa. Other threats that may not be detectable include global air pollution, climate change, and change in ecosystem structure and composition due the encroachment of invasive and alien species.

Direct destruction and fragmentation of ancient forests is a crucial threat to forests globally and can be seen on this map. However, threats that are not detectable, such as those listed above, should not be neglected.

Should the borders of the new protected areas be the same as the borders of intact forest landscapes as indicated on these maps, when it is decided to protect these areas?
Not always. Determining the borders of a protected area is a more complicated process than identifying intact forest landscapes. While laying out protected territories, one often must take into account the boundaries of river basins, traditional forest practices use by local people, natural values within disturbed or fragmented forests and other local conditions. The best possible limits of protected areas will often exceed the boundaries of remaining intact forests.

Is preserving intact forest landscapes enough to protect the biodiversity of the global forests?
No, it is not. In many regions of the earth, in particular in temperate forests, there are few or no intact forest landscapes left. Even in regions with remaining intact forest landscapes some types of ecosystems, threatened species and other natural valuables cannot be found in these intact forest landscapes. So, besides the intact forest protection, the protection of other types of high conservation value forests is needed, as well as improved management of forests to help protect biodiversity.