The soil food web of two beech forests (Fagus sylvatica) of contrasting humus type: stable isotope analysis of a macro- and a mesofauna-dominated community

The structure of the soil food web in two beech (Fagus sylvatica) forests, the Gottinger Wald and the Selling forest (Northern Germany), was investiga ted using variations in tissue N-15 concentrations of animal species or tax a. The Gottinger Wald is located on a limestone plateau and characterized b y mull humus with high macrofauna activity, particularly of Lumbricidae, Di plopoda and Isopoda. In contrast, the Selling forest is located on a sandst one mountain range and characterized by moder humus. The soil fauna of this forest is dominated by mesofauna, particularly by Collembola, Enchytraeida e and Oribatida. In June 1995 soil fauna was sampled using heat extraction. Three soil layers were analysed at each of the sites. N-15/N-14 ratios of bulk material increased strongly with soil depth in both forests. This also applied to the water-soluble fraction at the Gottinger Wald, but not at th e Selling. Generally, the water-soluble fraction was more enriched in N-15 than the bulk materials. For most animals studied N-15/N-14 ratios varied l ittle with soil depth. In both forests soil animals could be classified eit her as saprophages, including microphytophages, or predators. On average, t he delta(15)N of predatory taxa (Chilopoda, Araneida, Gamasina, Staphylinid ae) exceeded that of saprophagous or microphytophagous taxa (Lumbricidae, I sopoda, Diplopoda, Collembola, Oribatida, Enchytraeidae) by 4.4 and 3.9 par ts per thousand for the Gottinger Wald and the Selling, respectively. We as sume that most of the saprophagous or microphytophagous taxa studied consis t of primary and secondary decomposers and hypothesize that predators prey more on secondary than primary decomposers. Generally, average delta(15)N v alues differed little be-tween saprophagous (Lumbricidae, Diplopoda, Isopod a) and microphytophagous taxa (Collembola, Oribatida). The variations in de lta(15)N values of species within these taxa consistently exceeded the vari ation between them, indicating that the species of each of these taxa form a continuum from primary to secondary decomposers. Also, variations in delt a(15)N values within predatory taxa in most cases exceeded that between tax a excluding top predators like Sorer. We conclude that using higher taxonom ic units in soil food web analysis is problematic and in general not consis tent with nature. Higher taxonomic units may only be useful for depicting v ery general trophic groupings such as predators or microbi-detritivores.