Degree Name

Doctor of Philosophy (PhD)


Biological Sciences - faculty of Science


An understanding of the demography and population dynamics of rare species, and the mechanisms maintaining this rarity, is vital if we are to manage threatened species effectively. However, in long-lived plants this is no easy task, due to indeterminate growth and complex stage transitions in the lifecycle and cryptic life history stages such as soil-stored seed banks. In this study, I used a combination of classical demography, experimentation and matrix transition modelling to investigate commonness and rarity in obligate seeding species from the plant genus Persoonia. I made comparisons between two closely related species pairs, Persoonia lanceolata (common) and P. glaucescens (rare), and also Persoonia mollis subsp. nectens (common) and P. mollis subsp. maxima (rare), and included a third rare species, Persoonia bargoensis. If we are to predict and mitigate the effects of changing fire regimes on obligate-seeding vegetation, we need to understand the response of seed banks to fire. I quantified the dynamics of the cryptic seed bank stage by conducting a seed burial experiment where I monitored in situ seed survival and germination in both inter- and post-fire periods. I found high levels of seed dormancy and low levels of seed mortality in the absence of fire in both common and rare species. Germination only occurred in sites that burned in late 2001, and was significantly higher in the medium intensity fire than for either the low intensity or partial-burn fires. None of these seedlings survived to October 2002, possibly as a result of well below average winter rainfall for the region. Fire effects were patchy at small scales regardless of intensity, but became more uniform across the larger site scale as intensity increased. How seed banks respond to multiple fires is also poorly understood, yet this information is important if we wish to assess the ability of obligate seeding species to persist through such disturbance regimes. I quantified the response of the single seed bank of Persoonia mollis subsp. nectens to two closely spaced fires (1997 and 2001) by comparing the magnitudes of the post-fire germination responses and by using Ripley's L function to examine differences in the spatial pattern of the seedlings in each cohort at 3.5 years post-fire. There was strong germination from the seed bank after both fires (473 and 287 individuals at 3.5 years post-fire) indicating the ability of viable seeds to persist through at least two fires, although there were significantly fewer seedlings following the second fire. The spatial pattern of germination in both seedling cohorts was significantly clustered at small scales around deceased adults, suggesting limited dispersal. The second fire also resulted in tighter clustering around deceased adults and in fewer seedlings dispersed away from adults. My results suggest that Persoonia mollis at this site has a persistent seed bank that can buffer the population against unfavourable fire frequencies and post-fire conditions, and may allow it to retain higher levels of genetic diversity in the form of seeds from previous cohorts of adult plants. Matrix transition models that allow the calculation of the population growth rate, � and identification of the vital rates most influential in directing � are a useful tool for comparing the viability of different populations and examining potential management options. I collected demographic data on survival, growth and fecundity for the five Persoonia species in 2001-02 (a normal year) and 2002-03 (a drought year) and constructed size-based transition matrices for each species in each year. In all species � was not significantly different from 1 in the normal year, indicating that populations were stable under these environmental conditions. In the drought year however, all species and sites displayed a trend of reduced � but this reduction was only significant in the rare species P. glaucescens and P. bargoensis. Elasticity analysis revealed that individuals surviving and remaining in the same size class (stasis) were having the strongest influence on � in the normal year. In the drought year, however, it was survival in the seed bank that became most important in the rare species P. glaucescens and P. bargoensis, highlighting that an understanding of seed bank dynamics is essential for management. An analysis of the way in which variation in the vital rates influenced the observed change in � in the drought year (life table response experiments) revealed that changes in stasis and growth driven primarily by increased mortality contributed most to the change in �. Comparisons between the rare and common species within each year showed that the rare species are less able to buffer the effects of drought by maintaining consistent survival. This result suggests that management to reduce mortality in the rare species during low rainfall or drought periods may be an effective conservation approach. Because fire is an important mechanism that controls the dynamics of species in fireprone habitats, it is important to have an understanding of the most appropriate fire regimes to use in management. I constructed stochastic fire models to examine the relationship between fire interval and fire frequency and how this affects extinction probabilities and population size over the next 100 years. These models chose randomly from matrices for normal and drought years, with post-fire matrices constructed to simulate the change in population dynamics in the years immediately after fire. The model for regular fire-return intervals of between 3 and 20 years showed that return intervals of less than 8-10 years resulted in extinction of all populations. I then examined the effect of different combinations of inter-fire interval typical of those reported in the literature at a frequency of 10 fires in 100 years. I found that variation in fire interval length of alternating long and short intervals produced the most optimistic outcomes for both the rare and common species. Variable fire-return intervals have also been recommended as most suitable for maintaining a range of life history types in fire prone habitats. Regular return intervals produced the most pessimistic outcomes at this frequency. Interestingly, rare species performed more poorly than their common congeners in both the regular and variable fire return intervals. This study has provided valuable new information on the threatened Persoonia species in the Sydney region. By making comparisons of common and rare species using the matrix modelling approach, I was able to highlight differences in the life histories of the common and rare species and subsequently suggest management approaches. These differences were primarily due to the fortuitous occurrence of an infrequent drought year within the study period. This reinforces the need for longer-term studies in order to document the natural levels of life history variation in perennial plant populations. I also discuss the limitations of the current study and suggest areas where further research and data collection will help to refine the current findings.

02Whole.pdf (2463 kB)