Degree Name

Doctor of Philosophy


Department of Biological Sciences


Several models have been proposed to describe the main processes involved in the evolution or maintenance of hybrid zones. These differ in the type (genetic or environmental) and amount of selection acting on hybrids. Hybridizing species are often associated with specific habitats and empirical studies have revealed that the processes involved in maintaining hybrid zones are often complex and involve genotypic and environmental interactions. Short-lived species are often used to study hybrid fitness, because fitness can be assessed over the whole life-cycle within typical study periods. Hence, long-lived plants have been under-represented in hybrid zone studies.

The life histories of two Banksia species, Banksia robur and B. oblongifolia, provide an excellent opportunity to study the effects of hybridization in longlived, re-sprouting plants. The hybrid zones contain the products of past hybridization as modified by subsequent natural selection. However, the extent of hybridization and selection after a single recruitment event is unknown. I took advantage a fire in the Sydney region in January 2002, which burned a number of swamps and woodlands releasing seed from the Banksia canopy.

I developed genetic markers to determine the genetic structure of mature plants and seedlings within the hybrid zones and to detect hybrids. I developed eight primer pairs for Banksia microsatellite markers and genotyped individuals from B. oblongifolia, B. robur and B. paludosa populations, as the latter species potentially hybridized with the other two species. All eight loci were polymorphic, with the exception of two loci in B. robur. Numbers of alleles ranged from 1 to 13 and observed average heterozygosity ranged from 0.000 to 0.833. No linkage disequilibrium (P < 0.01) was observed between any pair of loci. At least four loci completely distinguished B. robur from B. oblongifolia and three separated B. paludosa from B. oblongifolia. Seven of these primers amplified DNA from at least two of three other local Banksia species.

To determine whether genetic data could both separate the taxa and predict morphology, I assessed B. robur and B. oblongifolia plants and seedlings using seven microsatellite markers and seven morphological characters. Pure stands of B. oblongifolia and B. robur mature plants formed genetically and morphologically identifiable groups. However, hybrid plants were not always intermediate. Morphology separated the two taxa grown in pots and hybrids had the same morphology as parentals. In contrast, seedling morphology in the field was highly variable. Greater phenotypic variation among field seedlings, due to drought conditions, particularly affected B. oblongifolia seedlings as they were growing in the dryer soils than B. robur seedlings. Therefore, B. oblongifolia seedlings had relatively fewer leaves making the two species morphologically more similar in the field than the two species in pots.

A range of hybrid genotypes was detected among seedlings. Continuing hybridization and introgression was revealed by the composition of the genotypes among seedlings and their maternal plants. The proportion of hybrids appeared to be consistent between generations (6.6% vs 6.2%).

Performance of the two species and their hybrids was assessed in the different environments of the hybrid zones, using indicators of fitness. Germination, survival, growth and herbivore damage were all monitored and assessed. Three methods were used to assess relative seedling performance. A transplant experiment was set up within the hybrid zones; transects of emerging field seedlings were monitored across the hybrid zones; and seedlings from the hybrid zones were raised in pots. Mature plants were also measured to determine whether hybrid plants produced as much seed and recovered as well as the parental species after fire.

Hybrid seedlings and plants were found in all three habitat types within the hybrid zones and, on average, germinated, survived and grew as well as parental species. Most hybrids were either similar to, or intermediate between, the two parental species in performance. Individual hybrids varied greatly however, and there was some evidence for hybrid superiority and also inferiority in a few seedlings. Whether these prove to be novel hybrid phenotypes depends on how they continue to develop.

There was no evidence for endogenous (intrinsic or genetic) selection against hybrid seedlings in hybrid zones or in pots. However, there was some direct and indirect evidence for selection against immigrant alleles in B. oblongifolia habitat (exogenous or environmental selection). There was less evidence for selection against B. oblongifolia and hybrid seed or seedlings in B. robur habitat. No selection was observed in intermediate habitat and each of the genotypic groups (B. robur, B. oblongifolia and hybrids) germinated, survived and grew equally well.

The Banksia hybrid zones appear to be maintained by gene flow assisted by common pollinators, adaptation of the two species to different environments and selection against immigrants within at least one parental habitat. Hybridization was restricted by the very narrow bands of intermediate habitat where the two species co-exist, partial temporal isolation through separation in peak flowering periods and infrequent recruitment events.