Numerical characterization of forest soils using biological and biochemical properties

Summary

To find if surface soils could be grouped by their biological and biochemical properties, soil samples (0–5 cm) were collected at 4-week intervals for 56 weeks from 48 woods in and around the English Lake District, and pH, loss-on-ignition (LOI), moisture content, oxygen uptake, and cellulase and phosphatase activities were measured. Results expressed on a loss-on-ignition basis were more informative than those on an oven-dry basis. In a principal component analysis of each property over the 14 samplings, the first component values represent “smoothed” between-plot differences; other components identify plots which behave differently from the majority at certain times. Analysis of variance showed very highly significant differences between plots for all the properties. pH and loss-on-ignition showed the smallest, but significant, differences between samplings. On a loss-on-ignition basis, the only significant correlations between first component values, and between plot means, were phosphatase with oxygen uptake and cellulase with pH. It is concluded that no individual physiological property can be used as a measure of soil bioactivity, which supports the conclusions of other authors. None of the principal component analyses of individual or combined properties showed any evidence of the existence of distinct clusters of plots. On a loss-on-ignition basis, a priori groups based on (1) pH < 3.8, (2) pH 3.8 – 5.0 and (3) pH > 5.0, showed no significant difference in moisture content. However, oxygen uptake was significantly lower in (1) than in (2). Cellulase activity was significantly greater in (3) than in (1) and (2). Phosphatase activity was significantly lower in (1) than in (2), and there appeared to be a peak at pH 3.8 – 5.0.