Optimal sampling design and minimal effort for soil charcoal analyses considering the soil type and forest history

Abstract

Soil charcoal analysis is of particular interest for reconstructing and interpreting past forest landscapes. However, whether soil charcoal spectra are representative of past forest communities or not remains unclear. Here we sampled three types of soils from two ancient forests in North France (six sites in total), using a single 1.5 × 1.5 m pit vs. several auger cores regularly distributed over a 1,000 m2 area (i.e. stand scale). Soil charcoals were extracted to compute specific anthracomasses (SA) and, for one of the two ancient forests studied, to determine taxonomic composition. We compared the two sampling methods (pit vs. auger) according to SA distribution within a pit and between auger cores using Moran’s I index. To determine the minimal sampling effort according to taxonomic richness, we used rarefaction curves. Except in the upper horizons of the podzol pits, within-pits SA distribution showed no spatial structure as a plausible effect of soil disturbance and biological activity. At the stand scale, between-augers SA distribution did not show spatial pattern. The pit method retrieved more species than the auger one, but the sampling effort was sometimes insufficient to be representative and has to be adapted according to the anthracomass. A minimal sampling effort of 500–600 charcoals was required to reach a full picture of the soil charcoal assemblage, corresponding to a volume of 30–60 litres depending on the sampling method. Historical land use and site management, together with the type of soil, should be taken into account when designing a sampling strategy for soil charcoals. The sampling design and effort should be adapted to the goal of the study, the biological activity of the soils and the expected density of charcoals.