Figure 1.
Simple schematic highlighting the five principle steps of our simulation model.
Table 1.
Generalised logistic regression model used to predict the proportion of hollow-bearing trees in each 10 cm DBH (diameter at breast height) cohort.
Table 2.
Linear regression model used to perform a sensitivity analysis of the mean number of hollow-bearing trees ha−1 (ln (x+1) transformed) perpetuated in urban greenspace over 300 years.
Figure 2.
Frequency distribution of median tree diameter cohorts for tree stands (all species) in nature reserves (open bars) and urban greenspace (solid bars).
Figure 3.
Simulations predicting the relative number of hollow-bearing trees ha−1 (mean ±95% prediction interval) over 300 years under existing management practices in nature reserves (1) and urban greenspace (2).
Figure 4.
The predicted relative number of hollow-bearing trees ha−1 (mean ±95 prediction intervals) in urban greenspace over 300 years for a range of values for variables with the greatest relative effect sizes derived from a sensitivity analysis.
Variables include: the maximum standing life of trees (A); the number of seedlings ha−1 (B); and the rate of hollow formation (represented by the coefficient for the probability of hollow occurrence; C). Predicted thresholds under existing management practices are provided for reference (solid circles).
Figure 5.
The predicted relative mean number of hollow-bearing trees ha−1 in urban greenspace over 300 years for a combination of values for: the maximum standing life of trees and the number of seedlings ha−1 (A); and the maximum standing life of trees and the rate of hollow formation (represented by the coefficient for the probability of hollow occurrence; B).
Figure 6.
The predicted relative mean number of hollow-bearing trees ha−1 over 300 years under a series of alternative urban tree management scenarios (dashed lines).
Simulated scenarios include: increasing the standing life of trees only up to 500 years; increasing the number of seedlings only up to 60 ha−1; accelerating hollow formation only by 62% above the observed mean rate (as represented by the coefficient for the probability of hollow occurrence); and a combined management approach (i.e. our recommended management proposal), which manipulates all three variables simultaneously. Scenarios under existing management practices are provided for reference by solid black lines for nature reserves (1) and urban greenspace (2).
Table 3.
Summarised values for each variable used to parameterise our simulation model under existing management practices for nature reserves and urban greenspace.