Soil preservation and experimental treatments
In April 2019, we collected soil samples from three habitats: a grassland (N 45.194° E 5.776°), a broadleaved forest (N 45.196° E 5.774°), and a vegetated river bank (N 45.195° E 5.780°). To allow DNA extraction immediately after sampling, all sites were within 400 m from the Laboratoire d’Écologie Alpine (LECA) in Grenoble, France. We established five plots within each habitat and the minimum distance between nearby plots was about 20 m. Within each plot, we collected four soil samples (with a minimum distance of one meter) at a depth of 0–20 cm and then pooled them together, for a total of five pooled samples per habitat (approx. 200 g each pooled sample). Soil litter was not included in the samples. Pooled samples (15 in total) were homogenized; subsequently, from each of them we took five subsamples of 15 g of soil (total: 75 subsamples; Fig. 1).
The five soil subsamples of each pooled sample were subjected to five different treatments: 1) eDNA was extracted immediately after sampling (within 1 h; treatment hereafter referred to as “control”); 2) samples were preserved at room temperature (21-23°C) and eDNA was extracted 6 h after sampling; 3) samples were inserted in sterile 50-mL falcon tubes and preserved at 4°C and eDNA was extracted three days after sampling; 4) samples were inserted in hermetic, sterile boxes with 20 g of silica gel immediately after sampling, then stored at room temperature, and eDNA was extracted 21 days after sampling; 5) samples were inserted in hermetic, sterile boxes with 20 g of silica gel 6h after sampling, then stored at room temperature, and eDNA was extracted 21 days after sampling.