Soil physicochemical properties for profiles of up to 1 m from savannah soils, Lopé National Park, Gabon, 2022-2023

Sampling regime
Samples were collected in 7 spatial blocks within the Lopé National Park, Gabon (0.156-0.192°S, 11.574-11.635°E) during two field campaigns in 2022 and 2023, respectively. Within each block, there were two savannahs with either frequent (annual) fire recurrence, or low fire recurrence (no more than 1 fire in past 20 years). Within each savannah, complete soil profiles were collected at three slope positions: Summit, Backslope, or Toeslope. At each of these slope positions, three profiles (spaced 10 m from each other) were collected with depth increments of 0-5 cm, 5-15 cm, 15-30 cm, 30-50 cm, 50-75 cm, and 75-100 cm. Samples from the three profiles were combined to give a composite sample for each fire frequency, slope position and depth. Data gaps exist where soil profiles could not be collected to 1 m depth owing to more shallow soil depth.

Sample processing
Samples were collected using a 5-cm diameter, 30 cm long corer (AMS soil corer). Intact cores were retrieved, separated by depth increment in the field, and transported to the research station in Lopé National Park in zip-lock plastic bags. The fresh weight was determined either on the day of sampling or the following day, and samples were stored in a ventilated place in open plastic bags to partially dry. All samples were sieved to 2 mm at the research station in a partially dried state. Stones and roots that did not pass the 2-mm sieve were removed and their mass recorded. Sub-samples of stones and all root material were collected to determine root dry mass and stone density for correction of total C stocks and bulk density. Complete soil samples for depth up to 50 cm, and representative sub-samples (around 50%) of larger deeper soil samples were transported under import licence to Stirling in zip-locked bags. In Stirling, all samples were oven dried (60°C, 48 hours) and re-weighed. Small subsamples (obtained by use of sample splitters to avoid inadvertent bias) were pulverised in a ball mill.

Analysis
C and N content (by mass) determined by elemental analysis (Flash SMART, ThermoFisher Scientific, Germany). Elemental analyser: %C and %N were calibrated against three standards of known concentrations. All data reported are deemed high quality based on application of calibration and consistency within location sites and across sampling depths. A total of 15 data gaps result from insufficient soil depth at 9 out of 42 locations.