Data from "Towards a biogeography of Amazonian soil microarthropods"

This dataset supports the findings in:Pequeno et al. (2023) Towards a biogeography of Amazonian soil microarthropods. Journal of Natural History. DOI: https://dx.doi.org/10.1080/00222933.2023.2268785<br>The study explored the roles of environmental factors vs. geographic location in driving the genus-level community composition of soil oribatid mites (Acari: Oribatida) in Amazonia. Mites were sampled in 2000 and 2001 across 111 sites distributed between two landscapes on opposite margins of the Amazon River: Reserva Ducke (Manaus, Western Brazilian Amazonia) and Alter do Chão (Santarém, Eastern Brazilian Amazonia).<br>Reserva Ducke (2.95°57’S, 59.93°56’W) is covered by wet forest encompassing extensive soil texture variation (from pure sand to 80% clay) to the north of the Amazon River near its confluence with the Negro River. There, sampling sites were regularly distributed over 10 × 10 km. Each sampling site consisted of one 250 m long transect with constant altitude, along which 20 litter–soil volumes(3.5 cm length ×3.5 cm width ×5 cm depth) were sampled. Each four consecutive volumes were placed in the same plastic container to reduce the processing load, resulting in a​ ​total of 5 compound samples. This material was kept cool in plastic containers and taken to the laboratory, where soil invertebrates were extracted with a​ ​Berlese-Tullgren apparatus into glass vials with 5% formaldehyde solution. Invertebrates were extracted for one week with daily increases in temperature (from 28 to 45°C).​ Then, oribatid mites were sorted from 100% of the extracted volume under a stereoscopic microscope.Alter do Chão (Santarém, Brazil; 2°3’S, 55°00’W) is a natural mosaic of dry forest fragments in a savanna matrix south of the Amazon River near its confluence with the Tapajos River. There, 54 sampling sites were spread over 15 × 12 km, 38 in the savanna and 16 in separate dry forest fragments. All sites were on relatively flat ground. Each site consisted of four parallel transects 250 m long, where 40 standardised volumes (3.5 cm length ×3.5 cm width ×5 cm depth) of the litter–soil interface were sampled. Soil processing followed as before​, except that oribatid mites were sorted from 50% of the total extracted volume, so that sampling effort per sampling site was standardised between landscapes.​ In both areas, oribatid mites were identified to genus level and counted.For each sampling site, litter dry mass was measured from the processed soil cores after fauna extraction, and soil clay content (particles &lt; 0.0002 mm) was measured from a separate set of six 5 cm deep soil cores collected along each transect (one each 50 m), using the hydrometer method. Both variables were averaged for each sampling site. Vegetation type was assigned based on in situ observation as savanna, dry forest, or wet forest.<br>The file "oribatids_amazonia.csv" contains a data table with 111 rows (sampling sites) and 146 columns (variables), as follows: sampling landscape, latitude (decimal degrees), longitude (decimal degrees), vegetation type, litter dry mass (g), soil clay content (%), and individual counts of genera in alphabetical order (columns 7 to 146).