Data from: Fast recovery of soil macrofauna in regenerating forests of the Amazon

This dataset includes information on macroinvertebrate abundance (broad taxonomic groups) and land-use history for 85 study sites of the eastern Amazon (in the states of Maranhão and Pará). The sampling took place annually between 2010 and 2016, at the end of the rainy season (April to August). The climate of the region is rainy tropical, with mean annual temperature of 27 ° C and annual rainfall of 2100-2300 mm. The most common types of soils are Oxisols and Argisols. The study sites were covered by ombrophilous forests both primary and at different successional stages.

Macrofauna sampling followed the Tropical Soil Biology and Fertility protocol (Anderson and Ingram 1993). At each study site, 3 to 8 samples were collected at each 20 m along a linear transect or two perpendicular linear transects in a cross shape. Each sample contained the litter and the top 10 cm of soil of 25 × 25 cm squares. All invertebrates larger than 2 mm were sorted by hand from samples and classified into broad taxonomic groups.

Land-use history was obtained from the interpretation of Landsat time series imagery. At each study site, we identified past events of forest clearing through the visualization of Landsat 4, 5, 7 and 8 images (available at http://earthexplorer.usgs.gov) acquired between 1984 and 2016. Clearing events were detected from  changes in forest patterns identified in true color image composites (combining red, green and blue bands) and images classified with the normalized difference vegetation index (NDVI). The number images used varied with year and location due to availability and cloud cover, but all sites were covered by at least one image per year. Overall, we analyzed 1508 images. The moment of forest clearance was not always observed but could always be dated with year-level accurately inferred from vegetation contrasts of sequential images. Field estimates of forest regeneration age and interviews of local people were used to validate satellite imagery interpretation.