From observational studies on succession in the field (Allen 1991) and experimental studies in two-species (e.g. Fitter 1977) and multi-species (e.g. Grime et al. 1987) mixtures, it is argued that AM fungi can alter the competitive balance between species and have a considerable impact on community structure. For example, Allen & Allen (1984) demonstrated an increase in the competitive ability of Agropyron smithui and Bouteloua gracilis in the presence of AM fungi when grown in interspecific competition with the nonmycorrhizal chenopod Salsola kali. Similarly, in a study using a de Wit replacement design to investigate the effect of AM fungi on competitive interactions between S. kali and the mycorrhizal grass Agropyron dasystachyum, Benjamin & Allen (1987) found that in the presence of AM fungi, the grass dominated in terms of biomass, but that in the absence of mycorrhizal colonisation, the chenopod became dominant. Evidence from a range of other experimental studies, in which one species is strongly responsive to the mycorrhizal symbiosis while the other is less mycotrophic or non-mycotrophic, generally indicates that AM fungi influence the outcome of competition when neighbouring species have differential responses to mycorrhizal fungi (Allen & Allen 1990). For example, Crush (1974) and Hall (1978) found that AM fungi improved the competitive ability of legumes when grown with grasses. Furthermore, Fitter (1977) found that AM colonization reduced the growth of Lolium perenne grown in interspecific competition with Holcus lanatus, while Hartnett et al. (1993) showed that the presence of mycorrhizal fungi greatly increased the competitive success of Andropogon gerardii in mixtures of the mycotrophic A. gerardii and predominantly non-mycotrophic Elymrus canadensis. In some of these studies the evidence for competition comes from simple comparisons of the performance of plants in monocultures and in mixtures. In others replacement series experiments have been used to calculate relative crowding coefficients (de Wit 1960) which are used to quantify the impact of mycorrhizal colonization on the competitive process. From a comparison of the effects of mycorrhizal colonization on a range of these coefficients (Allen & Allen 1990) it is clear that the addition of mycorrhizal inoculum did not generally reverse the competitive outcome, but did confer a varying advantage in terms of increased biomass to the mycorrhizal neighbour. Moreover, the response to mycorrhizal colonization in the monocultures and mixtures of a replacement series were not necessarily correlated. Unfortunately, the growth of plants in a replacement series in which monocultures and mixtures are grown at a single overall density does not allow the impact of AM fungi on intraspecific and interspecific competition to be distinguished. In a recent study published in the Journal of Ecology to investigate the impact of AM fungi on two cooccurring and positively mycotrophic grass species, Holcus lanatus and Dactylis glomerata, West (1996) grew the plants with and without AM inoculation at a range of densities in both monoculture and mixtures. Essentially the experimental design involved growing the plants in monocultures at densities of 1, 2, 4, 8 and 16 seeds per pot for each species, plus all paired combinations of the densities of each species. The influence of AM inoculation on the competitive ability of species i in relation to species j was then quantified by calculating the aggressivity index (Snaydon 1991) for each species combination at each density