1. Understanding how chronic anthropogenic disturbances are distributed in space leads to more effective conservation and management practices. This study provides a large-scale overview of how the Caatinga dry forest is potentially altered by chronic anthropogenic disturbance. In particular, we investigate how the intensity of the potential disturbance varies (i) among geographic regions, (ii) among fragment sizes, and (iii) as a function of distance to the fragment edge.

2. A Chronic Anthropogenic Disturbance Index (CDI) was derived to represent five anthropogenic disturbance vectors: human population, infrastructure, grazing, logging, and fire. CDI was spatialized for the whole Caatinga landscape (826 411 km2) and posteriorly averaged for ca. 47 100 remaining fragments.

3. Chronic anthropogenic disturbance was unevenly distributed across the Caatinga landscape, with more conserved sites on the western and southern regions. Larger fragments were, on average, as disturbed as smaller ones. However, smaller fragments varied from highly disturbed to very well preserved depending on the regional context. Strong edge effects were recognized for the chronic anthropogenic disturbance index and its five vectors.

4. Policy implications. Although half of the original Caatinga vegetation remains, most remnant fragments potentially suffer from chronic anthropogenic disturbances. This suggests that a large portion of the Caatinga can be currently threatened, and its biodiversity and natural resources eroded by processes that are less evident than habitat loss. Proper management practices and expansion of the existing conservation area network are needed to halt the Caatinga biodiversity decline.

The Chronic Anthropogenic Disturbance Index (CDI) surface represent the potential disturbance effect of anthropogenic disturbance vectors: human population, infrastructure, grazing, logging, and fire (For detailed description of variables and your weights used in each vector see Table S1 - Supplementary Material of Antongiovanni et al. (2020). 

The five anthropogenic disturbance vectors are:
Human population: was composed by human population density, distance to settlements, and distance to urban localities.
Infrastructure: was composed by distance to paved roads, distance to unpaved roads, distance to transposition channels, distance to wind turbines, distance to railroads, and distance to energy transmission lines.
Grazing: was composed by ovine density, caprine density and bovine density.
Logging: was composed by distance to charcoal-based industries.
Fire: was composed by fire density 

All 14 primary surfaces of disturbance (raster format) were expressed as distance or density and all disturbance variables surfaces were standardized by their maximum value (i.e., max = 100). The projection of the spatial data was Albers Equal Area Conic and the pixel size for all raster surfaces was 90 meters. All the spatial analyses were performed in ArcGis 10.1.

The relative importance of each primary variable was defined by assigning it a relative weight (wi), varying from 0 to 1 see Table S1 - Supplementary Material of Antongiovanni et al. (2020).

The CDI surface was calculated as the weighed sum of the 14 primary variables considering their relative importance, standardized by the maximal weighted sum recorded for in the landscape, including the anthropogenic matrix. Furthermore, in order to understand the spatial heterogeneity of different disturbance components, we also calculated the standardized weighted sum for the five disturbance vectors: human population, infrastructure, grazing, logging, and fire. The pixel values in these five surfaces were standardized according to the maximum value of each vector. 

All surfaces names are the same as described in the article and in supplementary material by Antongiovanni (2020) published in J. Apllied Ecology. The detailed description of the data source and the derivation of each vector and the CDI is also described in detail in the article and in the supplementary material by Antongiovanni (2020).