| Summary: | Experiments and a critical review of the literature of the technique of soil microbiology have shown that the agar plate method for the estimation of soil microbes may lead to severe errors. (1)The number of micro-organisms developing from a soil innoculum to colonies on agar media, and which are counted and used to estimate the soil population, do not always all develop. Of these organisms which can reproduce or grow on the medium provided, a proportion is not represented if the colonies are too crowded. The relationship between amount of soil innoculum and the number of colonies developing from it, is not then linear. When the development of some colonies is suppressed in this way, microbial populations estimated from the plate counts may be severely underestimated. The late developing colonies in particular are suppressed when the density of colonies on plates is high. The suppression of the development of microbial types into colonies is differential, and so changes in the degree of crowding of colonies on agar plates may bring about changes in the proportional representation of specific groups of microbes. The proportional representation of actinomycetes for example, is depressed at high densities, and yellow chromogens may be depressed or proportionately increased in different populations. It is shown that both the quantitative and qualitative effects of plating density can be attributed to microbial interaction, and that these effects are very significant within the conventional range of plating densities 70-200 colonies per 9cm. diamter plate. It is shown that there is a general relationship between the proportional representation of microbial types and the estimated total population of soils in the experimental results reported by a number of investigators. Many of the differences between rhizospheres and soil populations which have been claimed by investigators, and which show this relationship can be interpreted as plating density artifacts and it is pointed out that until many of the rhizosphere experiments which have been reported have been confirmed, making due allowance for or excluding the density effect, the results must be suspect. Statistical methods have been used to show differential rhizosphere effects (in the presence of the density effect) between three fen plants growing in peat, and between banana varieties. There is a correlation between the rhizosphere flora of the banana varieties examined and resistance to Panama disease. Within the genetic pairs Congo - Gros Michel, and Silk Fig - Guindy, the susceptible variety supports a larger microflora on its roots than the resistant. The power of the varieties to support a varied microbial flora is correlated with the size of population supported on root systems in the field. Antibacterial and antifungal substances were isolated from the rhizomes of a number of banana varieties, but no correlation with disease resistance could be found.
|
|---|
