DATASET: Abundance and diversity of Arecaceae in 100 20 m x 50 m plots along a 5-km transect at the RESEX Riozinho do Anfrísio, Eastern Amazon, Brazil

The present dataset resulted from a survey that was conducted within the federally-managed Riozinho do Anfrísio Extractive Reserve (hereafter, RESEX Riozinho), a 737,088-hectare natural reserve limited in the East by the Iriri River, a large clear-water tributary of the Xingu River in the state of Pará, Brazil. The survey was realised in May 2019, at the end of the rainy season. Palm trees were counted on 100 adjacent plots measuring 20 m x 50 m along a 5 km-long transect previously marked for mammal surveys, totalling a 10 ha cover area. Distances were measured using the portable measuring tool HipChain®, with every 50 m marked with flagging tape and mapped with a Garmin® 78S handheld GPS receiver. The transect was slowly censused on foot at ~1 km/h by a minimum of two experienced observers. All adult palms within ten meters on either side of the transect were recorded and identified to species level whenever possible using species descriptions. Their positions along the transect were then noted to the nearest 50 meters. The presence or absence of <i>igarapés</i>, shallow, narrow and slow streams typical of Amazonian environments, was recorded for each plot, generating a binary variable, as those are generally not visible from satellite imagery.Coordinates representing each 20 m x 50 m plot were later used to extracted data from the digital elevation model (DEM) database TOPODATA, from the Brazilian Spatial Research Institute (INPE), a country-wide resource which refined data from the Shuttle Radar Topography Mission (SRTM) and offers images with resolution of 30 m, containing measures such as altitude, slope, relief, exposure and terrain shape (Fig. 2). We used HydroSHEDS data to calculate the distance from each transect sub-unit to the nearest perennial stream as registered by the SRTM, which did not record recently formed <i>igarapés</i>. As a proxy to measure forest density, we extracted the normalized difference vegetation index (NDVI) for the coordinates representing each of the 100 plots, using satellite imagery from Cbers 4A . The calculations of NDVI, and topographical and hydrological variables, were conducted in QGIS 3.34 “Prizren”.