A matrix population model analysis for the tropical tree, Araucaria cunninghamii

The temporal stability of estimates of demographic behaviour for a sample population of the tropical rainforest gymnosperm Araucaria cunninghamii was analysed using a stage-classified (Lefkovitch) matrix model. Previously published life-table analyses, which used only 2 years of held data (1975–77), were compared with life-table analysis based on a further 5 years of field data for growth and survival (1977–82). The new analysis calculated the finite rate of natural increase, λ, as 1.009 ± 0.077. This value is closer to the theoretical value of 1.0 for a stable population than is the value calculated on the basis of only 2 years of field data (1.021 ± 0.093). However, the two estimates are not significantly different from each other or from a value of 1.0. Despite the similarity of estimates for λ, the life-table analysis based on 5 years of data showed some marked differences from the earlier analysis. Revised growth and survival rates imply a slower progression to reproductive maturity and shorter reproductive life for A. cunninghamii. The predicted stable stage distribution (SSD) showed a better fit to the initial stage distribution (ISD), and bulges in the frequency distribution for large trees can be explained by the varying rates at which individuals grow from one stage to the next. The relative contributions of survival, growth and fecundity to the observed population growth rate (elasticity) remained similar for both analyses. Survival rate contributes most significantly to the achieved population growth rate, and this appears typical of long-lived plant species. However, the pattern of elasticities across size-classes changed for the new analysis and identified different critical life history stages from those recorded previously. Elasticity is proposed as a measure for classifying species using the competitive (C), stress-tolerant (S) and ruderal (R) model of plant strategies.