The functioning of ecological systems can be studied at small spatiotemporal scales in controlled laboratory microcosms. In particular, the microcosm method has been used as a tool to understand the functioning of decomposer food webs (Scheu 2002; Huhta 2007). Although the experimental approach has limitations such as limited community composition and restricted mobility of animals (Kampichler et al. 2001), idealized model systems help to reduce the natural variability and exclude variables which are considered beyond the scope of the research. The dynamics of decomposer food webs and nutrients in such systems are fairly well known. Decomposer population dynamics in laboratory microcosms containing no auto­trophs, and hence no carbon input, consists of one growth phase and a subsequent decline phase (Nieminen 2008b). The length of the population growth phase depends on the initial population size (Nieminen 2002). For example, the biomass of fungal-feeding nematodes increased exponentially for 3.5 weeks in organic forest soil, then crashed and remained low for 16 weeks (Nieminen 2008b). In a laboratory experiment using larger pieces of forest floor, nematode populations crashed after 3 weeks, then collembolan and enchytraeid populations crashed (Huhta et al. 1983). In such a system lacking carbon input and nutrient uptake by plants, inorganic N accumulates in soil (Nieminen 2008b).

In such severely carbon limited systems, addition of organic carbon available to microbes increases microbial biomass, and reduces the N concentration in soil (Sparling and Williams 1986; Schmidt et al. 1997; Dunn et al. 2006). In an early experiment, glucose increased both the biomass of fluorescein diacetate active mycelium and yeasts as well as fungal-feeding nematodes in pine microcosms (Baath et al. 1978). However, the soil properties were quite unnatural, for example the pH was high (above 6.8), and algae thrived in microcosms but enchytraeids were lacking (Baath et al. 1978). In coniferous forest soil, the pH and moisture usually limit yeast and algal growth. Extra carbon input is also reflected in the biomasses of higher trophic levels of the soil decomposer food web. Addition of carboxymethyl cellulose to microcosms containing mineral and organic soil, needle litter and a Scots pine seedling increased the biomass of both saprotrophic fungi and hyphal-feeding nematodes (Nieminen and Setala 2001). Although the biomass of filamentous fungi increases after cellulose addition, addition of labile carbon such as sucrose can enhance the growth of early successional (r-strategist) microbes such as bacteria (Moore-Kucera and Dick 2008; Nottingham et al. 2009) and Zygomy — cota fungi (Hanson et al. 2008). Nieminen (2010) found that a sucrose addition equalling 100 kg C ha-1 maintained a stable nematode population for one growing season, and an increasing enchytraeid population. Sucrose addition did not alter net N mineralization rates, indicating that N mineralization by increased animal popula­tions exactly balanced the N immobilization in microbial biomass. When the carbon addition rate is increased above 100 kg C ha-1 in Norway spruce forest soil, soil animals, which have orders of magnitude lower growth rates than microbes, cannot consume all the extra microbial biomass in one growing season, and as a conse­quence, N is immobilized in microbial biomass.