Predicting the dynamics of a native Araucaria forest using a distance-independent individual tree-growth model

Abstract

In recent decades, native Araucaria forests in Brazil have become fragmented due to the conversion of forest to agricultural lands and commercial tree plantations. Consequently, the forest dynamics in this forest type have been poorly investigated, as most fragments are poorly structured in terms of tree size and diversity. We developed a distance-independent individual tree-growth model to simulate the forest dynamics in a native Araucaria forest located predominantly in southern Brazil. The data were derived from 25 contiguous plots (1 ha) established in a protected area left undisturbed for the past 70 years. The plots were measured at 3-year intervals from their establishment in 2002. All trees above a 10-cm diameter at breast height were tagged, identified as to species, and measured. Because this forest type comprises hundreds of tree species, we clustered them into six ecological groups: understory, subcanopy, upper canopy shade-tolerant, upper canopy light-demanding, pioneer, and emergent. The diameter increment, survival, and recruitment sub-models were fitted for each species group, and parameters were implemented in a simulation software to project the forest dynamics. The growth model was validated using independent data collected from another research area of the same forest type. To simulate the forest dynamics, we projected the species group and stand basal areas for 50 years under three different stand-density conditions: low, average, and high. Emergent species tended to grow in basal area, irrespective of the forest density conditions. Conversely, shade-tolerant species tended to decline over the years. Under low-density conditions, the model showed a growth tendency for the stand basal area, while under average-density conditions, forest growth tended to stabilize within 30 years. Under high-density conditions, the model indicated a decline in the stand basal area from the onset of the simulation, suggesting that under these conditions, the forest has already reached its maximum-stock capacity. The model validation using independent data indicated close agreement between the observed and estimated values, suggesting the model is consistent in projecting species-group and stand growth. The methodology used in this study for developing the growth model should be tested in other species-rich forests.