Soil and rainforest composition in Tasmania: correlations of soil characteristics with canopy composition and growth rates in Nothofagus cunninghamii associations

Abstract

The relationships between soil chemistry and canopy composition and growth rate were examined in several Nothofagus cunninghamii rainforest associations. There was considerable variation in the concentrations of some of the soil nutrients, with c. 40-fold differences in total phosphorus among surface samples (5–15 cm depth) and 170-fold differences among deep samples (50–60 cm depth). Principal components analysis with rotated axes indicated that surface extractable K, total N and loss of ignition (LOI) contributed most to Component 1. The factors contributing most to Component 2 were total exchangeable bases, extractable Mg and LOI of the deep samples and surface extractable Ca. Total phosphorus (surface and deep), pH and surface C: N contributed most to Component 3. The soils of N. cunninghamii-dominated forests had significantly higher pH and total phosphorus than mixed rainforests (rainforest without a clear dominant species) and a lower C: N ratio than soils of both mixed rainforests and P. aspleniifolius-dominated rainforests (P < 0.05). However, no significant differences were recorded in any soil parameter between the latter two forest types. Factor 3 of the PCA was positively correlated with the abundance of N. cunninghamii and negatively correlated with abundance of P. aspleniifolius and E. lucida (P < 0.05). In addition, the growth rate of N. cunninghamii was positively correlated with total phosphorus after removal of high-altitude sites (≥700 m a.s.l.). These results, together with previous data on comparative growth rates, suggest that phosphorus has a significant influence on the canopy composition of these rainforests via its effect on the growth rate of N. cunninghamii. Concentrations of both total and available phosphorus were very low on some sites, overlapping the range of values recorded in button grass plains, heaths and eucalypt forests from other studies in Tasmania. More comprehensive data are required to allow detailed comparisons among these vegetation formations. However, the results are consistent with Jackson’s (1968, 1983) hypothesis that the absence of rainforest from some low-nutrient soils may be influenced more by fire frequency (via the interactions among soil nutrients, vegetation and fire) than directly by soil nutrients.