Mesofauna and macrofauna densities at species/group level from 2008 to 2020 in three regions in Germany

Global biodiversity loss is threatening ecosystem functioning and human well-being. Arthropods above the ground have substantially decreased in abundance and diversity during the last 15-20 years. However, changes in belowground biodiversity, particularly in forests, received little attention. Here, we analysed a comprehensive dataset of soil-living meso- and macrofauna in forests differing in land-use intensity within the framework of the open research platform ‘Biodiversity Exploratories’. Abundance of soil animal species was analysed at three-year intervals, covering 12 years from 2008 to 2020. Neither density and species richness nor gamma diversity of both soil meso- and macrofauna declined, suggesting contrasting dynamics of biodiversity above and below the ground. Soil mesofauna density and diversity varied significantly between years within regions, with the variations being closely related to soil moisture of the previous winter months and during sampling. While the stability of mesofauna and of some macrofauna taxa was strongly correlated with asynchrony of species-fluctuations, and in part with effective diversity, overall neutral or positive variance ratios suggested that most species fluctuated synchronously. These synchronous fluctuations were likely due to variations in abiotic conditions such as soil moisture and presumably overprinted biotic drivers of stability. Stability was not directly related to forest management, although for mesofauna, it differed between forest types within regions. While documenting an astounding resilience of soil animals in temperate forests to the ongoing biodiversity decline, our findings help to better understand temporal patterns of soil fauna density and diversity and the drivers of their community stability.

Methods
Samples were taken in three regions in Germany (Swabian Alb, Schorfheide Chorin, and Hainich-Dün), spring (April to June) from 2008 to 2020 in intervals of three years, resulting in 5 sampling points. Samples were taken from 5 m × 5 m subplots located within the 100 m × 100 m gridplots as described in Erdmann et al. (2012) and Klarner et al. (2014). Briefly, the four forest types were replicated four times in each region, resulting in 48 sampled forests. In each forest, one soil core of 5 cm Ø was taken for the extraction of mesofauna and one soil core of 20 cm Ø for the extraction of macrofauna. We separated the litter layer (variable thickness) and the top 5 cm of the soil underneath for extracting meso- and macrofauna by heat (Macfadyen 1961). Individuals from the two layers were pooled for statistical analyses. Lumbricidae were sampled from an area of 0.25 m² from each forest using a combination of hand sorting and extraction with mustard solution. First, the litter layer was removed and checked manually for earthworm specimens. Then, a mustard solution consisting of 100 g mustard powder (Semen Sinapis plv.; CAELO, Cesar & Loretz GmBH, 40721 Hilden, Germany) dissolved in 10 l water was applied to the soil in 2 steps with initially 5 liters and another 5 liters after 15 min (Eisenhauer et al. 2008). Emerging Lumbricidae were hand-collected for a total of 30 min. All animals were stored in 70% ethanol until identification. Mesofauna (Oribatida, Collembola, and Mesostigmata) and macrofauna taxa (Isopoda, Diplopoda, Lumbricidae, and Chilopoda) were identified to species level using appropriate keys (Eason 1964; Hopkin 2007; Karg 1989; Karg 1993; Krantz & Ainscough 1990; Oliver & Meechan 1993; Schubart 1934; Sims & Gerard 1985; Weigmann 2006). Araneae and Coleoptera were only sorted at the taxon level and were not considered in the analyses of diversity.

Precipitation for each year and sampling site was derived from the RADOLAN (Radar Online Adjustment) product of the German Weather Service (Deutscher Wetterdienst), which provides hourly radar-based precipitation estimates for Germany adjusted to rain gauge data on a 1 km² grid (Kreklow et al. 2019). Temperature was measured with environmental sensors installed at 2 m above the ground at all sites (Fischer et al. 2010). We calculated mean values for temperature and precipitation in winter (December to February) and spring (March to May). We also used the silvicultural management intensity indicator (SMI), including a risk and density component, and accounting for tree species, tree age, and aboveground living biomass (Schall & Ammer, 2013; 2023); it was taken from the Biodiversity Exploratories database (BExIS). Microbial biomass in leaf litter and soil was assessed by measuring the maximum initial respiratory response (MIRR; mg O2 g^−1 h^−1) after glucose addition (SIR method; Anderson & Domsch 1978) in an automated O2 micro-compensation apparatus (Scheu 1992). Glucose (80 and 10 mg g^−1 dry weight for litter and soil, respectively) was added as an aqueous solution to approximately 1 g of leaf litter material (Beck et al. 1997).