Bacteria and microbial-feeders modify the performance of a decomposer fungus Jouni K. Nieminen * , Heikki Seta Èla È University of Jyva Èskyla È, Department of Biological and Environmental Science, P.O. Box 35, FIN-40351 Jyva Èskyla È, Finland Received 12 September 2000; received in revised form 18 January 2001; accepted 4 April 2001 Abstract Westudiedwhetherthepresenceofabacteriumcanaffectthefunctioningorstabilityofsimplefungal-baseddecomposerfoodchains.We constructed microcosms with 60 g washed mineral soil and four different food webs: 1) a fungus Cladosporium herbarum) alone; 2) a fungusandafungal-feedingnematodeAphelenchoides sp.); 3)afungusandabacteriumEscherichia coli);and4)afungus,abacterium,a fungal- and a bacterial-feeding nematode Aphelenchoides sp.and Acrobeloides tricornus).Glucosewassuppliedasthesolecarbonsource. One replicate set of microcosms was kept at 228C for the sixth and seventh week as an experimental disturbance. The microcosms were destructivelysampled10timesforfoodwebeffectsnon-disturbedcontrols)andthreetimesforfreezingtreatment.Fungalbiomassgrowing alone was less ef®cient i.e. it had higher respiration rate per unit biomass) than the other food webs. This did not, however, result in the greatest C loss in the long term, because the fungal biomass declined when growing alone. Presence of the bacterium increased the decomposition of dissolved organic carbon as much as the presence of the fungal-feeder, but their effects were not additive. Ammonium nitrogen concentration of the medium was higher in systems containing bacteria. Nematode populations and the bacterial biomass were not sensitive to the disturbance of freezing. Although the freezing disturbance decreased total fungal biomass in the absence of the bacterial energy channel, dissolved organic carbon decomposition was only incidentally retarded. In spite of their distinct role in the system functioning, the bacteria did not signi®cantly alter the stability properties of the system. Thus, our results contradict some recent food- web based decomposition models, which predict that the effects of parallel energy channels on decomposition are additive. We discuss nutrientlimitationandthedependenceoffungalpropertiesonthefood-webcon®gurationasexplanationsforthecontradictionoftheoryand data.Wedidnot®ndevidencesupportingahypothesisthatanadditionalenergychannelshouldstabilize¯owthroughafoodweb,andhence decomposition. This is explained by high resilience of the reference system with a fungus alone, and by redundancy, i.e. the fact that the fungal-feeding nematode induced similar changes in the fungus than the bacterium. q 2001 Elsevier Science Ltd. All rights reserved. Keywords: Microbial food web; Decomposition; Freezing; Energy channel; Stability 1. Introduction The relation of food-web organization to the functional and dynamical ecosystem properties is of current interest, largely due to the potential environmental and economic consequences of declining biodiversity. The ecological questions are not, however, new McNaughton, 1988, and references therein): over 40 years ago Elton 1958) argued that species richness explains the balance of communities. Based on modelling the relation between species richness, food-webconnectanceandinteractionstrength,May1972) concluded, contrary to expectations, that as a mathematical generality, diversity decreases stability. The relevance of mathematical generality to natural reality, however, can be quite justly questioned Berendse, 1994). Food-web compartments have been offered as an explanation for the existence of diverse systems May 1972; Moore et al., 1996). Coleman et al. 1983) divided rhizosphere activities into hypothetical `fast' bacterial- and `slow' fungal-based components, and Moore and Hunt 1988) produced quantitative evidence of resource compartmentation of soil food webs. Carbon ¯ows through the decomposer food web along two distinctive routes: bacterial and fungal energy channels Hunt et al., 1987; Wardle and Yeates, 1993), the dominance of which varies between ecosystem types Ingham et al., 1989). The bacterial and fungal channels do not remain separate throughout the web, but are merged by general- ist predators at higher trophic positions Moore and Hunt, 1988). Soil Biology & Biochemistry 33 2001) 1703±1712 0038-0717/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved. PII:S0038-071701)00093-1 www.elsevier.com/locate/soilbio * Corresponding author. Tel.: 1358-14-2601211; fax: 1358-14- 2602321. E-mail address: jkniemin@cc.jyu.® J.K. Nieminen).