Search

Banksia microphylla leaf fossils and Banksieaeidites zanthus pollen are newly described from late Eocene lignite of the Zanthus-11 borehole, drilled east of Norseman in Western Australia. The leaf fossils are the first known in Banksia to show extreme narrowness (

The pollination biology of Anigozanthos humilis (Haemodoraceae) was studied within a Banksia woodland reserve using a combination of field techniques and genetic analysis. Motion-triggered cameras were deployed on 25 flowering plants to identify visitors to flowers, quantify visitation rates and assess visitor behaviour. Entire A. humilis plants were caged to exclude potential floral visitors with six treatments: (i) bird and honey possum exclusion, allowing access by insects; (ii) bird and insect exclusion, allowing access by honey possums; (iii) total animal exclusion; (iv) open, allowing access by birds, insects and honey possums; (v) hand-pollination with cross-pollen; and (vi) hand-pollination with self-pollen. Open pollinated seed were genotyped to assess mating system parameters including outcrossing rate. From 23424 h of surveillance with camera traps, 109 visits were recorded, 106 (97%) of which were western spinebills (Acanthorhynchus superciliosus), with three visits by brown honeyeaters (Lichmera indistincta). We recorded an average of 8 (±4) visits per plant, equivalent to one visit every eleven days over the 2016 flowering season (mean = 63 days). Mean seed set per fruit was negligible (0.2–0.3 seed per fruit) for all pollination treatments except open- and hand-pollination with cross pollen. Mean seed set per fruit for hand cross-pollinated flowers (37.1 seed per fruit) was 37 times that of open pollinated flowers (1.0 seed per fruit). Outcrossing rate estimates were not significantly different to one. Our results indicate that at least for the study population and season, A. humilis was largely dependent on western spinebills for delivery of outcross pollen and resulting seed set, but visitation rates were low and pollen limitation was severe. Given the known sensitivity of western spinebills to habitat fragmentation, our results suggest flow on effects that may negatively impact on reproduction, and means that A. humilis may be particularly vulnerable to environmental changes that impact on bird pollinators.

In the South-west Australian Floristic Region, timing of rainfall is critical for successful seedling establishment, as is surviving the first year’s summer drought for population persistence. Predictions of a warmer, drier future, therefore, threaten the persistence of obligate seeding species. Here, we investigate the drought tolerance of two co-occurring Banksia (Proteaceae) species by withholding water in pots to different extents of soil drying. Seed was collected from high- and low-rainfall populations, to test for niche differentiation in water-use strategies at the species level, as well as population differentiation. On the basis of a more negative leaf water potential at minimal levels of stomatal conductance and quantum yield, B. coccinea was considered to be more drought tolerant than B. baxteri. This was supported at the anatomical level according to xylem-vessel attributes, with a higher estimated collapse pressure suggesting that B. coccinea is less vulnerable to xylem cavitation. Population contrasts were observed mainly for B. baxteri, with a lower leaf-expansion increment rate in the low-rainfall population providing for drought avoidance, which was reflected in a higher rate of survival than with the high-rainfall population in which 87.5% of plants showed complete leaf senescence. The implications of species differences in water-use strategies are that community dynamics may start to shift as the climate changes. Importantly, this shift may be population dependent. A systematic understanding of adaptive capacity will help inform the choice of population for use in revegetation programs, which may lead to increased resilience and persistence in the face of environmental change. The results of the present study suggest that should declines in B. baxteri populations be noted, revegetating with seed collected from the low-rainfall population may help improve the chances of this species surviving into the future.

Given the predicted changes in rainfall patterns for many Mediterranean climate regions, identifying seed tolerance to moisture stress in the earliest phase of plant development is an important consideration for species conservation, management and restoration. Here, we used polyethylene glycol (PEG 8000) to induce plant water deficit similar to drought stress in a field situation. Seeds of four Western Australia Banksia R.Br. (Proteaceae) species were incubated at seven levels of moisture potential (0 to −1.5 MPa) and three constant temperatures (10°C, 15°C and 20°C). In the absence of moisture stress, germination was uniformly high, but increasing drought stress led to reduced and delayed germination in all species. Overall, the threshold moisture potential value for a significant decline, and delay, in germination was –0.25 MPa. Results suggested that one species (B. coccinea) is likely to be most vulnerable to germination failure under predicted changes in rainfall patterns, whereas another (B. media) is likely to be less vulnerable. There was significant variation in population response to drought stress. However, this variation could not be explained by rainfall across species distributions. We discuss the PEG approach for assessing seed sensitivity to moisture stress, particularly in the context of shifting rainfall under climate change.

Intervals between fires are critical for the persistence of obligate-seeding shrubs, and are often used in planning fires for fuel reduction and biodiversity conservation in fire-prone ecosystems worldwide. Yet information about the trajectories of reproductive performance for such species is limited and information is often qualitative. To test existing assumptions about reproductive maturity periods for eight obligate-seeding shrubs (with both canopy and soil seedbanks) in foothill forests of south-eastern Australia, we used a chronosequence approach, with sites from 2 years to >40 years post-fire. Quantitative measurements of flowering and fruiting were used to fit models of reproductive response in relation to time-since-fire for each species. Inferred reproductive maturity for each species, based on modelled times to reach 80% of maximum flower production, varied from 5 to 18 years post-fire. For a subset of three species, models predicted 80% maximum seed production occurring 1–7 years later than flowering. Our results confirmed or extended assumptions about post-fire reproductive maturity for these species, and provided a basis for improved incorporation of plant life-history in ecological fire planning. We infer that increased fire frequency makes one of our study taxa, Banksia spinulosa var. cunninghamii (Sieber ex Rchb.) A.S.George, vulnerable to decline because of its long reproductive maturity period and serotinous seed storage.

Limited success of restoring framework banksia-woodland species has been attributed to the failure of seedlings to establish deep root systems before the onset of the summer drought. The present glasshouse study investigated how optimising nutrient application during nursery production may increase new-root production after outplanting. Two experimental streams were established to (1) optimise nutrient application rates during nursery production and (2) utilise nutrient-loading techniques to improve root production of Banksia menziesii R.Br., Banksia attenuata R.Br. and Eucalyptus todtiana F.Muell after outplanting. Optimal nutrient-application rates were determined by measuring plant growth and internal nutrient responses to eight application levels of slow-release fertiliser (0–18 kg m–3, nitrogen (N) : phosphorus (P) : potassium (K) = 17 : 1.6 : 8.7). Nutrient-loading treatments utilised seedlings that had been grown under common industry fertiliser conditions (3 kg m–3 native Osmocote, N : P : K = 17 : 1.6 : 8.7) supplied with ‘low’ or ‘high’ loading doses of liquid Thrive continuously over 6 weeks, immediately before outplanting. Seedlings from both experiments were then outplanted to 1-m-deep poly-pipe tubes containing habitat soil. After 12 weeks, plants were harvested and new-root production and shoot growth were measured. Optimal concentrations of slow-release fertiliser for maximum outplanting success as indicated by increased root investment (root : shoot ratio and new-root production) were 8–12 kg m–3 for all species. Nutrient loading increased N and P concentrations of plants by up to 80% and 127%, respectively, by luxury nutrient consumption, and after planting, nutrient-loaded seedlings produced 1.5-fold the biomass of conventionally fertilised seedlings, this being the result of greater root productivity. In conclusion, optimising nursery nutrient regimes for framework species may increase root-growth potential, assisting in improving plant establishment in restoration programs.

The soilborne plant pathogen Phytophthora cinnamomi is listed as one of the world’s 100 worst invasive alien species by the International Union for Conservation of Nature (IUCN). The impacts on native flora and fauna habitats have been identified as a key threatening process in Australia. Identifying and mapping diseased vegetation and the rate of spread of the disease is required for management; however, this is often difficult and costly. This study investigated the ability of using a time series of orthophotos (1953–2008) in combination with Landsat satellite imagery, including trend analysis, and GIS to identify the presence of vegetation impacted by P. cinnamomi at four sites in Banksia woodlands in Western Australia. Further, the historical extent and rate of spread of P. cinnamomi was assessed at one site between 1953 and 2008. Our assessment identified that three of the four sites were affected by P. cinnamomi, results that are consistent with on-ground surveys. Investigation of disease progression at one site found a large increase in the area impacted between 1974 and 1988 and the rate of spread was highest between 1953 and 1963 (1.286 m year−1) and lowest between 1997 and 2008 (0.526 m year−1). The techniques presented provide a cost-effective tool to monitor broad-scale vegetation dynamics over time for management of this plant pathogen.

Diverse Cenozoic (possibly latest Oligocene to mid–late Miocene) macrofossils from the Lightning Ridge opal fields are illustrated and discussed. Specimens identified to, or closely comparable with, extant taxa include ferns (Lygodium, Gleichenia and others), conifers now extinct in Australia (Dacrydium, Retrophyllum and Papuacedrus), Lauraceae (Cryptocarya/Cinnamomum), sclerophyllous Proteaceae (Banksia, Lomatia and Grevillea), Cunoniaceae/Elaeocarpaceae and Eucalyptus (and/or other Myrtaceae). Overall, at least four fern, three conifer and 30 angiosperm taxa are recognised. The climate supported many species with close relatives in wet Australasian habitats, including rainforests. However, a drier or more seasonal (?monsoonal) aspect is especially indicated by the presence of lobed leaves that resemble extant species of Brachychiton (Malvaceae), Erythrina (Fabaceae) and tribe Cercideae (Fabaceae). A degree of water stress is also suggested by the prevalence of narrow, toothed and/or deeply lobed angiosperm leaves.

In southern New South Wales, a suite of widespread plant species exhibit short-statured ‘dwarf’ growth forms on coastal headlands. It is unclear whether such populations are genetically distinct or whether dwarfism is a plastic response to the environment. We used four microsatellite markers to assess genetic differentiation among populations from coastal and inland sites for Banksia spinulosa var. spinulosa. We sampled plants from six locations, including from three ‘dwarf’ and three ‘normal’ populations. Mean levels of genetic diversity were slightly higher in the forest (Na = 7.07 ± 0.25; He = 0.80 ± 0.09) than on the coast (Na = 5.92 ± 0.70; He = 0.72 ± 0.10). In general, populations displayed genotypic diversity expected for outcrossed sexual reproduction, with 161 of 172 individuals displaying unique genotypes and mean values of FIS close to zero. However, we found evidence of at least limited clonal replication in each of four populations and, within one coastal population, 11 of 27 individuals displayed one of three replicated genotypes, implying that the effective population size may be considerably smaller than would be inferred from the number of plants at this site. Relative to studies with other Proteaceae, this set of populations showed low, although significant, levels of differentiation (global FST = 0.061; P

Phytophthora cinnamomi continues to cause devastating disease in Australian native vegetation and consequently the disease is listed by the Federal Government as a process that is threatening Australia’s biodiversity. Although several advances have been made in our understanding of how this soil-borne pathogen interacts with plants and of how we may tackle it in natural systems, our ability to control the disease is limited. The pathogen occurs widely across Australia but the severity of its impact is most evident within ecological communities of the south-west and south-east of the country. A regional impact summary for all states and territories shows the pathogen to be the cause of serious disease in numerous species, a significant number of which are rare and threatened. Many genera of endemic taxa have a high proportion of susceptible species including the iconic genera Banksia, Epacris and Xanthorrhoea. Long-term studies in Victoria have shown limited but probably unsustainable recovery of susceptible vegetation, given current management practices. Management of the disease in conservation reserves is reliant on hygiene, the use of chemicals and restriction of access, and has had only limited effectiveness and not provided complete control. The deleterious impacts of the disease on faunal habitat are reasonably well documented and demonstrate loss of individual animal species and changes in population structure and species abundance. Few plant species are known to be resistant to P. cinnamomi; however, investigations over several years have discovered the mechanisms by which some plants are able to survive infection, including the activation of defence-related genes and signalling pathways, the reinforcement of cell walls and accumulation of toxic metabolites. Manipulation of resistance and resistance-related mechanisms may provide avenues for protection against disease in otherwise susceptible species. Despite the advances made in Phytophthora research in Australia during the past 40 years, there is still much to be done to give land managers the resources to combat this disease. Recent State and Federal initiatives offer the prospect of a growing and broader awareness of the disease and its associated impacts. However, awareness must be translated into action as time is running out for the large number of susceptible, and potentially susceptible, species within vulnerable Australian ecological communities.

There is a history of discrepancy in the interpretation of floral symmetry in the genus Banksia L.f. The variation arises not just from the potential nuances of symmetry terminology, as variation is also found in the diagrams that describe it. This paper clarifies the issue of the orientation of the ovaries within a Banksia unit inflorescence. Dissection of 2400 flowers in situ from four Banksia species has revealed the Banksia unit inflorescence to be enantiomorphic (having left-right asymmetry) due to the orientation of the ovaries. The asymmetry of the ovary and its pronounced development of the anterior side in fruit formation means that fruits are also readily distinguished as having arisen from either a left- or right-handed flower.

We have studied the ecology and conservation requirements of Banksia species in the species-rich sandplains of south-western Australia for 25 years. Loss of habitat through land-clearing has had the greatest impact on their conservation status over the last 50 years. Ascertaining optimal conditions for conservation management in bushland requires detailed knowledge of the species under consideration, including demographic attributes, fire regime, growing conditions and interactions with other species. Where populations have been fragmented, seed production per plant has also fallen. The group most vulnerable to the vagaries of fire, disease, pests, weeds and climate change are the non-sprouters, rather than the resprouters, with population extinction so far confined to non-sprouting species. Recent short-interval fires (

This paper details some of the recent research findings concerning restoration needs of the Banksia woodland in Western Australia, including the importance of, and recent advances in, smoke-technology research. Research has enabled testing of a wide spectrum of restoration technologies that enhance plant replacement at sites via treatments of the topsoil seedbank, broadcast seed and seedlings. By the use of smoke technology, which in some systems produces a 48-fold increase in the total number of germinants and a 3-fold increase in the number of species at restoration sites, improved species replacement is a very real possibility in Banksia woodland. At the same time, some commonly employed practices in restoration are a cause for concern, including the application of a herbicide widely used to control a priority Banksia-woodland weed and fire-suppressing agents used to assist fire management. These findings may have broader implications for restoration programs.

The consideration of phreatophyte response to changes in water availability is important in identifying ecological water requirements in water-resource planning. Although much is known about water-source partitioning and intra- and interspecific variability in groundwater use by Banksia woodland species, little is known about the response of these species to groundwater draw-down. This paper describes a preliminary study into the use of xylem cavitation vulnerability as a measure of species response to reduced water availability. A response function and critical range in percentage loss of conductance is identified for four Banksia woodland overstorey species. Similarity in the vulnerability curves of B. attenuata R.Br. and B. menziesii R.Br. at low tensions supports the notion that they occupy a similar ecohydrological niche, as defined by their broad distributions relative to depth to groundwater. B. ilicifolia R.Br., however, as an obligate phreatophyte, has a range restricted to environments of higher water availability and shallower depth to groundwater and this is reflected in greater vulnerability to cavitation (relative to other Banksia) at lower tensions. The wetland tree Melaleuca preissiana Schauer generally expressed a greater vulnerability at any given xylem water potential (Ψx). This paper identifies the range in Ψx within which there is an elevated risk of tree mortality, and represents a first step towards quantifying the critical thresholds in the response of Banksia woodland species to reduced water availability.

The conservation of wild orchid populations may depend on the establishment of propagated orchids to field sites to help sustain depleted populations if natural recruitment is not successful. However, very little is known about biotic factors which influence the establishment of terrestrial orchid seedlings in natural habitats. An experiment was established to measure the survival of six orchid species during their first growing season following transplantation to a West Australian urban bushland with a Banksia and Eucalyptus canopy and understorey dominated either by weeds or native vegetation. Symbiotically germinated orchid seedlings raised in the laboratory for 5 months before planting were established in adjacent field sites with high or low weed cover. There was a gradual mortality of seedlings at field sites throughout the growing season, primarily owing to insect grazing, and this was not affected by the enclosure of seedlings by wire mesh or shade cloth. Overall rates of survival varied from 49% for Microtis media R.Br., a species capable of growing in disturbed habitats, to 21% for Caladenia arenicola Hopper & A.P.Brown, the most common native orchid at these sites. However, not all surviving seedlings produced a tuber, so their expected rate of survival after the next dry season was reduced further. The factors having the greatest impact on seedling survival were site aspect (slope and canopy cover), weed cover and orchid species respectively. Orchid seedling survival was not well correlated with the presence of existing orchids of the same species at the same sites or the presence of compatible fungi in soil at these sites (simultaneously measured by orchid seed baiting).

We tested the ability to predict plant (and hence population) age for three fire-sensitive obligate-seeder Proteaceae species (Banksia ericifolia, Banksia marginata and Petrophile pulchella) in the heath and woodland vegetation of the Sydney region. To do this we sampled the number of growth whorls, as well as other growth characteristics (stem girth and height measurements, and canopy area and volume estimates), in areas of known time since last fire (TSLF). The average number of growth whorls was a very good predictor of plant age for both Banksia species (R2 = 98%, 99%), but this needed to be corrected for linear underestimation in P. pulchella (R2 = 92%). This technique could successfully be applied to these species in similar habitats across a large spatial scale, and so this information can be used to determine the age of a population in areas of unknown TSLF. A sample size of 15 plants was sufficient for accurate age estimates of all species; however, better estimates of TSLF for a particular plant community were obtained when estimates from two or more of the species were combined. We thus provide empirical evidence for the validity and accuracy of the growth-whorl technique for predicting plant age and hence TSLF. This information will assist in informing the development of appropriate management strategies for plants in relation to fire. Of the other growth characteristics studied, stem girth was the most reliable predictor; however, in general these other characteristics had wide confidence intervals on the predictions for sites greater than 10 years TSLF, owing to a non-linear relationship with age.

Plants pollinated predominantly by vertebrates are thought to have suites of floral traits (e.g.�colour, conspicuousness, odour) that favour either birds or mammals, with brightly coloured, conspicuous flowers associated with birds and drab, concealed flowers with non-flying mammals. This study examined two other floral traits, diel patterns of nectar production and pollen presentation (anthesis). It would be expected that these would be nocturnal in putatively mammal-pollinated plants and diurnal in bird-pollinated plants. In four Banksia and two Dryandra species, all known to be visited by honeyeater birds and small marsupials at one site in south-western Australia, there was no clear correspondence between visual cues and diel patterns of resource presentation. This lack of correlation between floral traits does not support the idea of specialised pollination syndromes, but rather is consistent with generalised pollination systems.

This study compares, for the first time, variation in estimates of susceptibility of native flora to Phytophthora cinnamomi Rands among four databases and proposes an estimate of the proportion of the flora of the South-West Botanical Province of Western Australia that is susceptible to the pathogen. Estimates of the susceptibility of south-western native flora to P. cinnamomi infection were obtained from databases for Banksia woodland of the Swan Coastal Plain, jarrah (Eucalyptus marginata Donn. ex Smith) forest, the Stirling Range National Park and Rare and Threatened Flora of Western Australia. For the woodland, forest and national park databases, hosts were naturally infected in uncontrolled diverse natural environments. In contrast, threatened flora were artificially inoculated in a shadehouse environment. Considerable variation occurred within taxonomic units, making occurrence within family and genus poor predictors of species susceptibility. Identification of intra-specific resistance suggests that P. cinnamomi could be having a strong selection pressure on some threatened flora at infested sites and the populations could shift to more resistant types. Similar estimates of the proportion of species susceptible to P. cinnamomi among the databases from the wide range of environments suggests that a realistic estimate of species susceptibility to P. cinnamomi infection in the south-western region has been obtained. The mean of 40% susceptible and 14% highly susceptible equates to 2284 and 800 species of the 5710 described plant species in the South-West Botanical Province susceptible and highly susceptible to P. cinnamomi, respectively. Such estimates are important for determining the cost of disease to conservation values and for prioritising disease importance and research priorities. P. cinnamomi in south-western Australia is an unparalleled example of an introduced pathogen with a wide host range causing immense irreversible damage to unique, diverse but mainly susceptible plant communities.

In Australia, numerous methods have been used to determine the time since the last fire at a given site. One method involves counting the number of annual growth nodes on Banksia spp. that are either killed by fire, or regenerate from surviving rootstocks, to determine above-ground plant age. Although a number of studies have used the Banksia node-count method to estimate plant and therefore site age, no published data assess the reliability of this method. This study attempted to determine the accuracy of the method, with shrub-form B. marginata individuals from five sites of different age, in a south-eastern Australian sand heath. A significant relationship was found between modal node (internode) counts and the known time since fire, with internode counts at four of the five sites accurate to within 1 year of actual site age up to 21 years (n > 50 individuals per site). The results suggest that the Banksia node-count method is a useful tool for determining site age up to 21 years. However, sample sizes need to be appropriate for the area sampled, given the potential error in counting nodes and the inherent site variability in age classes.

Banksia species are conspicuous in Australian heathlands and woodlands and provide a major source of nectar for a range of vertebrate and invertebrate pollinators. However, nectar production by Banksia is difficult to measure because inflorescences comprise large numbers (hundreds to thousands) of tiny flowers. Our aim was to quantify variation in nectar volume among B. ericifolia L.f. plants, but existing techniques were too time consuming or required destructive sampling of inflorescences. We estimated nectar volume by counting the number and estimating the size of nectar drops on a sample of flowers on each inflorescence. Our technique was fast (10 min per inflorescence) and a highly significant positive relationship was found between visual estimations and direct measures (capillary tubes and micropipettes) of nectar volume (r = 0.92; P ≤ 0.0001). This technique thus provides an alternative to current techniques for measuring nectar volume in plants that have large numbers of small flowers arranged in dense inflorescences. By using this technique, we found significant variation among days in nectar volume produced by each of five B. ericifolia plants (P ≤ 0.02). Furthermore, days of high and low nectar production were synchronised among plants. We concluded that the temporal variation found may be the result of fluctuations in environmental variables, rather than the age of an inflorescence. Indeed, nectar volume was found to be significantly correlated with a range of weather variables including daily maximum temperature (°C), precipitation (mm), wind speed (km h–1) and relative humidity (%). However, larger-scale and longer-term studies are needed to adequately examine any relationship between environmental variables and nectar production.

The regeneration strategies of three Banksia species in relation to fire were studied over 20 years in a mediterranean heathland-shrubland on the south coast of Western Australia. Banksia baueri and B. nutans are both bushes 1–2 m high, while B. baxteri is a shrub 4 m high. All three species regenerated only from seed released from the canopy seed bank after fire. They did not start to flower until 6 years after fire and seed set took even longer. Differences between the species in age-related intensity of flowering were related to the rate at which each species accumulated seed in the canopy. Even plants over 40 years old were still increasing their overall canopy seed bank or replacing seeds that had been released or were no longer viable. The vegetation studied appeared to be little affected by humans historically and to have burnt only at intervals of 30–60 years or more. Consequently, although all three species needed fire to regenerate, management of fire regimes needs to allow adequate intervals between fires for the replacement of their canopy seed banks. Indeed, all three Banksia species studied were extinguished from one area burnt twice at an interval of 9 years. Models developed with Banksia species from the northern sand plains of Western Australia, where fires appear more frequent, may need modification to be applicable to all south-coastal species.

Two species of banksia, studied on the south coast of Western Australia, flowered simultaneously and had floral characteristics often associated with vertebrate pollination. The pollen of both was carried by honeyeater birds and small mammals (honey possums). Despite this, differential exclusion of vertebrates and invertebrates from inflorescences indicated that Banksia attenuata set substantial quantities of seed when visited only by invertebrates, although seed set increased with the addition of vertebrate visitors. Banksia baxteri set much seed in the absence of any animal visitors and seed set increased with invertebrate, but not vertebrate, visitation. We suggest that these differences in pollination strategies reflect regeneration by B. baxteri solely from canopy seeds released after fire, whereas B. attenuata also regenerates from seeds released between fires, as well as from lignotubers and epicormic buds.

Despite the accumulating evidence that non-flying mammals are effective pollinators, further research is required to clarify how widespread this phenomenon is. The role of non-flying mammals as pollinators of four species of Banksia was investigated in north-eastern New South Wales. Nine species of non-flying mammals were captured amongst flowering Banksia and all carried variable amounts of Banksia pollen on their fur or in their faeces. Although not captured, feathertail gliders (Acrobates pygmaeus) were observed foraging at Banksia inflorescences. Squirrel gliders (Petaurus norfolcensis) visiting B. integrifoliaand pale field-rats (Rattus tunneyi) visiting B. ericifolia, carried substantial loads of pollen. Fur pollen loads for these species were of a magnitude similar to those of nectarivorous birds that were sampled closer to the time of foraging. Assessment of newly opened flowers indicated that considerable amounts of pollen were removed at night. The results of a pollinator exclusion experiment were inconclusive but B. ericifolia inflorescences exposed to nocturnal pollinators had consistently high fruit-set. This study lends additional support to the notion that pollination of Banksia by non-flying mammals is widespread.

Lomatia silaifolia (Smith) R.Br. (Proteaceae) is a common shrub in southeastern Australian bushland that generally flowers only once after each fire. However, little is known of the details of this post-fire flowering. Lomatia silaifolia has flowers well-spaced along its conflorescence axis, unlike many other Proteaceae species, thus allowing close examination of the influence of flower position, mate choice and flowering sequence on fruit production. We examined breeding system, flowering phenology and spatial patterns of fruit set in the species, after a fire in September 1992 at Bulli Tops, New South Wales. Flowering occurred in December–January, but only in the first and second summers after the fire. The species is partially self-compatible: only 25% of self-pollinated conflorescences initiated fruits compared with 100% of cross-pollinated conflorescences. Fewer flowers initiated fruits after hand self-pollination (3.0%) than after cross-pollination (35.4%). Self-pollinated flowers produced fewer viable seeds (22.2% viable) than cross-pollinated flowers (62.9% viable). Pollen tubes were found in 72% of the self-pollinated flowers examined, suggesting that there is no early stylar self-incompatibility in the species. There was a high level of herbivory on flowering and fruiting branches, with 69% of unbagged branches completely destroyed. This may significantly affect recruitment in the species, given the limited opportunities for reproduction in the post-fire environment. In this study, fruits were evenly distributed along the conflorescence axis unlike some other species in the Proteaceae with more compressed, spike-like flowering structures where fruits are typically concentrated in some parts of the conflorescence axis (e.g. Telopea and some Banksia spp.). No bird visitors to flowers were observed, but a variety of insect visitors were identified.

Effects of incursion of water and nutrients from agricultural land on adjoining native Banksia woodland were examined near Moora, Western Australia, by using a 100-m-wide belt transect from the paddock edge into virgin bush. Transect zones 0–20, 20–50 and 100–300 m inwards were designated as ‘hedge’, ‘transition’ and ‘inner bush’, respectively. Numbers of species of native woody taxa increased from 28 to 42 and plant densities increased from 2200 to 14 900 plants per hectare from hedge through transition to inner bush, while collective total standing biomass (above and below ground) of woody species decreased from 157 to 32 t dry matter per hectare. Data for nutrient contents of this total biomass indicated that total nutrient loads in respect of calcium, potassium, sulfur, phosphorus, magnesium, zinc, copper and sodium were closely related to relative amounts of biomass in the three zones, but hedge vegetation was appreciably enriched in nitrogen but depleted in manganese and chlorine compared to inner bush. Diversity indices indicated differential responses of species to agricultural influence. The dominant tree species, Banksia prionotes, represented 91, 87 and 58% of the total biomass of woody taxa in hedge, transition and inner bush, respectively. Growth ring analyses of basal trunk xylem indicated that this seeder species had recruited after a hot burn in 1963 coincident with land clearing. Since then, annual trunk-area increments in hedge trees have been consistently much greater than in transition and bush. Net annual productivities, dry matter allocation profiles and current leaf areas of B. prionotes showed that the average hedge tree weighed 167 kg total dry matter and was currently increasing at 11.9 kg per year and was in the process of achieving net dry matter gains of 420 g m–2 of leaf surface. Comparable figures for transition trees were 61 kg, 4.7 kg per year and 360 g m–2, for inner bush trees 12.7 kg, 1.1 kg per year and 230 g m–2. Results are related to recent studies on Banksia woodlands and general information on ecotones and their attributes.

The nectar sugar composition is presented for 147 species from 16 genera of South African and Australian Proteaceae. Patterns associated with flower age, different plants and populations, plant phylogeny and pollination have been examined. In addition to the usual three nectar sugars (sucrose, fructose and glucose), the nectar of Protea and Faurea contains the pentose sugar xylose at concentrations of up to 39% of total sugar. Xylose has not previously been reported from floral nectar and is absent from the nectar of Adenanthos, Banksia, Brabejum, Dryandra, Grevillea, Hakea, Lambertia, Leucospermum, Macadamia, Mimetes, Orothamnus, Paranomus, Stenocarpus and Telopea. Most genera and species have hexose-dominant nectar, but within the large genera Banksia, Grevillea, Leucospermum and Protea some of the seemingly more derived species have sucrose-dominant nectar. This interesting dichotomy of low versus high sucrose is of diagnostic value at the species level and indicative of phylogenetic relationships within the larger genera. At the generic level, the presence of xylose is a convincing synapomorphy for Protea and Faurea. Studies of physiological processes (e.g. enzyme activities) and ecological processes (e.g. pollination) may help to explain some of the conservative and taxonomically interesting nectar sugar patterns.

There has been a significant increase in the number of studies investigating plant breeding and mating systems over the past 10 years. The family Proteaceae, in particular, has dominated such research conducted in Australia. Thus it is now timely to present a critical review of the breeding and mating systems of the Australian Proteaceae. It is hoped that this will stimulate further research. The review covers key events between pollen deposition on stigmas through to fruit set. The genus Banksia, although not the most diverse of the family, has received a disproportionate amount of attention. It has featured in nine published studies of self-compatibility compared to 13 studies spanning the other 45 genera and has featured in eight genetic studies of the mating system compared to just two on other genera. Few studies have assessed the timing of stigma receptivity despite the intriguing situation in most Proteaceae of auto-deposition of self-pollen on or near stigmas at anthesis. Studies suggest that stigmas are not receptive until 0.5–4 days after anthesis. Banksia species appear to show low levels of self-compatibility although one subspecies shows high levels of selfing and evidence of selective fruit development. Self-compatibility may be more common in other genera, although a dearth of studies precludes generalisation. Assessment of mating systems indicates almost complete outcrossing for most species, lending support to the idea of selective fruit development. It is clear that many further studies of all topics are required but particularly across a wide range of genera because many have not been studied at all.

Interspecific and intergeneric pollen-tube growth was investigated using controlled hand pollinations of Banksia coccinea R.Br., 35 species of Banksia L.f., and three species of the related genus Dryandra R.Br. Currently, the relationship between B. coccinea and other species groups within Banksia is unclear. Some species supported no germination of B. coccinea pollen, some supported normal pollen-tube growth and others produced pollen-tube abnormalities including thickened walls, bulbous swellings, directionless growth, burst tips and branched tubes. Control of pollen-tube growth in the pistil was imposed in the pollen presenter and upper style. There was no significant reciprocal effect on pollination success in the lower style. The results of pollen-tube compatability in the lower style indicated that B. coccinea has a closer affinity to the section Oncostylis, than the section Banksia where it is currently placed. Intergeneric crosses of B. coccinea with Dryandra species resulted in some compatibility, with one cross having low numbers of pollen-tubes in the pollen presenter and upper style region. These results indicate a close relationship between Banksia and Dryandra, which are sister genera in the tribe Banksiae, family Proteaceae.

Banksia integrifolia L.f. has been in decline on calcareous sands of the Yanakie Isthmus, southern Victoria, since early 1980. Early studies indicated that the decline is associated with a particular soil condition possibly a nutritional imbalance involving Fe. However, in foliage samples collected from the three main soil types of the Isthmus, declining trees had similar concentrations of Fe but lower concentrations of Ca than healthy trees. Comparisons were made of seasonal variation in concentrations of macro- and micro-nutrients in foliage and litterfall from healthy trees (to minimise secondary changes associated with decline) within healthy and declining sites on the same soil type. On average, litterfall and the nutrient content of litterfall was greatest within the canopy area of B. integrifolia of the healthy stand. Banksias of the healthy stand also had greater concentrations of N, P, K and Na in fully-expanded leaves, resorbed greater proportions of phloem-mobile nutrients from senescent leaves and accumulated more Ca in senescent leaves. However, there was no evidence of nutritional imbalance in healthy trees within declining stands. It is argued that the lower foliar Ca in declining trees on three soil types and lower nutritional status of healthy trees within declining stands were due to lower productivity and lower water use and were therefore a result or an indication of decline rather than a cause.

Phytophthora cinnamomi Rands was isolated from either dead plants or soil at 46 disease centres in Banksia woodland at national parks and reserves on the Swan Coastal Plain. Phytophthora cryptogea Pethybridge & Lafferty was also isolated from roots of dead Acacia pulchella R.Br. in one disease centre infected with P. cinnamomi. Dead plants were infected with Armillaria luteobubalina Watling & Kile in four disease centres on the Spearwood Dune System, and these centres were excluded from further analysis. Phytophthora cinnamomi diseased areas ranged from 0.01 to 30 ha in size (mean 1.6 ± s.e. 0.7 ha). The total area infested for the 46 disease centres was 71.5 ha. Impact of P. cinnamomi was high in 17% of disease centres and low in 11% of disease centres. Age of plant death was a mixture of old and recent in 85% of disease centres. Mainly old deaths occurred in only 4% of disease centres. The proportion of species dying in infested areas varied between 10-64% (mean 28 ± s.e. 2%) and was positively correlated with impact type. It was found that infestation decreased species number; on average, there were seven fewer species in infested compared to non-infested areas. Four plant species associated with moist sandy sites tended to occur more frequently in centres of high impact than by chance alone. Occurrence of P. cinnamomi was related to soil association with soils of 60% of the disease centres belonging to the Bassendean or Southern River associations of the Bassendean Dune System. Sixteen percent of disease centres occurred in the Cannington, Guildford and Serpentine River associations of the Pinjarra Plain. No disease centres of P. cinnamomi were found on soils of the Speanvood and Quindalup Dune Systems. A water table was found within 3 m of the soil surface in 48% of the centres. Disturbance was associated with all disease centres. Firebreaks were associated with 72% of disease centres. Banksia woodland remnants on the Bassendean Dune System and the Pinjarra Plain are highly vulnerable to infection by P. cinnamomi and their conservation requires control of existing infestatinns and protection from introduction af the pathogen.

Estimates of the susceptibility of plant species in Banksia woodland to Phytophthora cinnamomi Rands were obtained by determining the incidence of plant death and frequency of isolation of the pathogen, among species occurring in 46 disease centres on the Swan Coastal Plain south of Perth, Western Australia. In the disease centres, dicotyledons outnumbered monocotyledons. About half of all species occurring in the disease centres were from four families of dicotyledons, with the largest number of species from the Myrtaceae, Proteaceae and Papilionaceae. The greatest number of species of monocotyledons were from the Anthericaceae and Cyperaceae. No deaths were recorded for 47% of species found in three or more disease centres. These species were mainly from the Cyperaceae, Haemodoraceae, Myrtaceae and Papilionaceae. The species that tended to die frequently in disease centres were mainly from the Papilionaceae, Proteaceae, Epacridaceae, Xanthorrhoeaceae and the Zamiaceae. Phytophthora cinnamomi was isolated from 26 of the 95 species occurring in three or more disease centres. For most species, the frequency of isolation of P. cinnamomi from recently dead plants was much less than the frequency of dead plants sampled. Isolation from plants was less frequent than from adjacent soil. The pathogen was isolated from recently dead plants or soil mainly for species of the Proteaceae, Myrtaceae, Papilionaceae, Dasypogonaceae, Iridaceae and Xanthorrhoeaceae. Cross-tabulation of species by incidence of plant death and isolation of P. cinnamomi from plant and soil, provided the opportunity to classify the response of plant species to infection by P. cinnamomi.

Banksia includes 38 fire-killed (seeders) and 20 resprouting species, and two species with contrasting ecotypes, in south-western Australia. There may be up to 12 seeders per 50 × 50 km grid cell in the southern sandplains and 12 resprouters in the northern sandplains. The patterns of distribution of species across soil type and eight climatic attributes is similar for both life forms, except that greater numbers of resprouting species occur at higher rainfalls and where there is greater seasonal spread of rainfall. Most seeders occur on white sands and rocky substrates, and resprouters occur on yellow sands and the more fertile lateritic soils. Nutrient requirements for both life forms appear to be similar. Resprouters are more widespread than seeders which suggests that resprouters show greater environmental tolerances. The distribution of grid cells containing each life form across soil types and eight climatic attributes is similar and any differences in climatic profile for all species in each category are considered biologically insignificant. Both life forms in section Abietinae are well represented in the climatically distinct southern and northern sandplains indicating no climatic preferences within the lineage. There are no consistent trends in environmental attributes from fire-killed to resprouting ecotypes of B. ashbyi E.G.Baker and B. violacea C.A. Gardner. Multiple-partitioning classification of the floristic data produced 10 groups varying greatly in geography, species richness, and proportion and endemism of each life form. The Lesueur (northern) district has four endemic seeders, six endemic resprouters and a mean of 10 resprouters per cell. The East Eyre (southern) district has five endemic seeders, no endemic resprouters and one resprouter per cell. Both groups have a mean growing season of 5 months. The relative aridities and fluctuations of present and past (Quaternary and late Tertiary) climates are invoked to explain the much higher proportion of resprouters in the northern than southern sandplains and the absence of seeders in the most marginal cells. The absence of endemic species yet high proportion of resprouters (73%) in the extreme south-western corner of the region might be explained by elimination of seeders through frequent burning by Aborigines in the late Quaternary. The increase in the proportion of fire-killed species along the south coast from 23% to 100% at the edge of the Nullarbor Plain also requires an explanation.

The rate and mechanism of the spread of Phytophthora cinnamomi in Banksia woodland of the Bassendean Dune system north of Perth, Western Australia was studied. Aerial photographs were used to measure the spread of disease in fronts longer than 5 km over 35 years. Fronts in upper slope positions moved downhill and uphill at 1.01 and 1.13 m year-1, respectively, while fronts in low-lying positions moved downgrade and upgrade (with and against the direction of water table flow) at 1.30 and 1.20 m year-1, respectively. Fronts in low-lying areas spread significantly faster than those on upper slopes. Excavations to 2 m depth of the root systems of 21 dying plants of Banksia attenuata R.Br., an overstorey co-dominant, revealed that 28% (3.9 roots/tree) of all first-order roots were infected. Assessment of the proportion of roots infected revealed a significant (P < 0.01) departure from uniform distribution of disease down the profile in both upper slope and drained flat sites. In 14 trees on dune slopes, 6-40 m above the aquifer, infection rate was higher than expected in roots lying at 21-40 cm depth, but lower than expected in roots below 1 m. Seventy three percent of infected roots lay in the top 40 cm of soil. In seven trees on a drained flat, 2-3 m above the aquifer, infection rate was high in the 21-60 cm horizon and also in roots below 1 m. Even though Phytophthora cinnamomi was active in the vicinity of the water table in trees on drained flats, the slow, steady spread of the disease in all landscape positions suggested that the primary mode of fungal invasion was through roots of susceptible vegetation. No evidence was found of accelerated spread of disease caused by the dispersal of zoospores.

Mating system parameters of the rare and endangered species Banksia brownii were estimated for two populations using both the mixed mating and effective selfing models. Estimates of outcrossing rate were similar in both populations for both models (mixed mating t Pop. 1=0.68, Pop. 2=0.75; effective selfing t Pop. 1=0.65, Pop. 2=0.73) and were among the lowest reported for undisturbed Banksia populations. Banksia brownii is killed by fire and the high level of selfing found may be associated with this trait. Multilocus and minimum variance mean i estimates were similar and the covariance of selfing with gene fixation (D) was not significantly different from zero indicating that populations were not structured and that most of the inbreeding was the result of self-fertilisation. The absence of structure was attributed to gene dispersal through pollen disperse by birds, and selection against inbred seed. It is suggested that several entire, large populations of this species together with habitat sufficient to support pollinators be reserved to conserve this species.

The morphology of the pollen presenter and pollen was studied in 79 taxa of Banksia using scanning electron microscopy. There were three major types of pollen presenter. Elongated and cylindrical structures were characteristic of the series Banksia, Crocinae, Tetragonae, Coccineae and some species of the Cyrtostylis and Abietinae. Short and cylindrical pollen presenters were found in the series Salicinae, Quercinae, Prostratae and some species of the Cyrtostylis. Short and ovoid pollen presenters were found in the series Grandes and Dryandroideae, and in the species B. brownii and B. ericifolia of the Spicigerae and B. pulchella of the Abietinae. The pollen-receptive stigma cells of all taxa were enclosed within a groove, which in the majority of species was obliquely terminal and longitudinal. Subterminal grooves, all of which were longitudinal, except that of B. speciosa which was located laterally, were observed in a few species. Pollen of all taxa had two germination pores, at opposite ends of a cylindrical, ovoid or crescent-shaped grain. There was a higher level of between species and between group consistency in pollen grain than in pollen presenter structure. The results have implications for both the interspecific fertility relationships and taxonomy of the genus.

Simulations were used to investigate the effect of season of fire on seedling emergence in Banksia ericifolia and B. serrata in the Sydney region, New South Wales. The simulations were based on models of soil-surface moisture in response to rainfall, seedling emergence response to soil moisture and post-fire seed release from fruits as determined by fire intensity, derived from field and laboratory studies. Fires were modelled on the first day of each calendar month for a 50-year period. Levels of post-fire seedling emergence were calculated using rainfall data from the Sydney Observatory for the corresponding period (1931-1980). Trends in seedling emergence as a function of month of fire were examined. Alternative sets of simulations were performed to assess the effect of variations in post-dispersal seed mortality, fire intensity and induced summer dormancy. In both species, mean emergence was affected by season of fire only when a 10% per month level of post-dispersal seed mortality was simulated (there was no fire-season effect at lower mortality levels). Highest predicted emergence occurred after summer fires and lowest emergence after winter fires. A reduction in rate of seed release (lower intensity fire) and induced seed dormancy in summer also had a minor effect with respect to fire-season in B. ericifolia. Reported levels of post-dispersal seed mortality in Banksia species are often high, and therefore, the simulations suggested that there will be an effect of fire season on seedling emergence. However, given the high level of year to year variation in seasonal rainfall in the Sydney region, fire-season effects are not predictable in the short term. In the longer term, the timing of fire relative to sequences of wet and dry years may be of equal importance to season of fire in its effect on populations of these species.

Seed bank dynamics of three co-occurring, non-sprouting Banksia species (B. baxteri, B. speciosa and B. coccinea) in patch-burnt scrub-heath (aged 10 and 21 years) were studied in the southern sandplains of Western Australia. In the younger plants, canopy seed storage was highest in B. coccinea. However, 21-year-old B. coccinea stored an order of magnitude fewer seeds than the other species and no more than young B. coccinea. Seed storage per year increased exponentially with plant age in B. speciosa and B. baxteri, whereas a quadratic function, peaking at 16 years, best described seed storage in B. Coccinea. Degree of serotiny was lowest in B. coccinea. Seed abortion did not vary between age cohorts but was highest in B. coccinea. Granivory ranged from 18 to 42% and was highest in the older plants and in B. baxteri. Cockatoos removed about 20% of cones in B. baxteri and B. coccinea and 10% in B. speciosa. Experimental cone removal accelerated follicle opening, especially in B. coccinea, irrespective of whether cones were placed on the ground (simulating cockatoo removal) or reattached to the plant (simulating plant death). Half the 21-year-old B. coccinea were dead and the remainder were considered senescent, as most branches showed dieback and cone fertility had fallen to 8%. Death of the other species was negligible, with cone fertility rising to a mean of 45% (B. speciosa) and 66% (B. baxteri). Interfire establishment was significant only in B. coccinea. The build-up of a viable seed bank with time occurred at a slower rate than for non-sprouting banksias in the northern sandplains.

Three Banksia spp. accounted for 80% of the aerial biomass of scrub-heath on deep sand. Reproductive structures contributed relatively more than leaves or stems to biomass of a non-sprouting Banksia, which also retained most of its dead leaves, than to the two resprouting species. Sclerophyllous hemicryptophytes accounted for only 1% of the aerial biomass. Litter contributed 19% of above-ground dead plus live phytomass and was poorly decomposed. The overall root f shoot ratio, to a depth of 2.5 m, was a high 2.35. Rootstocks, laterals and proteoid roots made major contributions to below-ground phytomass even though they were confined to the uppermost 15 cm of soil. Despite an exponential decline in root mass with depth, the gradient was less than in related ecosystems due to the abundance of well developed tap roots and vertical laterals (sinkers) to a depth of at least 5 m.

Three factors that potentially influence fruit set (breeding system, inadequate cross-pollination and variable floral display) were examined for two species of Banksia. Firstly, self-pollination and autogamy treatments failed to produce any fruit in B. spinulosa and produced very few fruit in B. paludosa. This suggests that cross pollen is required to produce the level of fruit set observed in the field. Secondly, when plants were given abundant cross pollen, fruit set was not altered in B. spinulosa but was in B. paludosa. Although the number of follicles per inflorescence was was not different for open- pollinated (control) and manipulated B. paludosa inflorescences, the latter produced twice the proportion of inflorescences with follicles produced by control plants and twice the total number of follicles per plant. Thirdly, larger floral display in open-pollinated B. spinulosa plants was associated with a greater production of fruit. However, larger floral display was not associated with a greater reproductive output in B. paludosa, and we suggest that recurrent pollen-limitation in this species may be responsible for this result.

We studied the post-fire vegetation development of a low open woodland dominated by Banksia attenuata and B. menziesii near Perth, Western Australia. Two similar stands burned in autumn and spring displayed different regeneration patterns, with seedling regeneration occurring only in the autumn burn area. Vegetative regrowth was more rapid and post-fire species numbers were higher in the spring burn area. Introduced annuals increased significantly in the autumn fire area. Longer-term vegetation development was studied in a series of stands ranging in age since last fire from 1 to >44 years. Species richness was greatest in the 5-year-old stand, and many shrub species were most abundant 2-5 years after fire. Non-native annuals were found only in stands less than 5 years old since last fire. Dominance by the shrub Eremaea pauciflora increased with stand age, although shrub structure and total biomass did not vary greatly except in the oldest stand studied. The proportion of total shrub biomass accounted for by leaves declined with stand age. Both the two major Banksia species had mixed size structures with seedlings present in all stands, indicating that neither is dependent on fire for recruitment. The results indicate that while autumn burns promote seedling regeneration they may also increase invasion by non-natives, and spring burning may be preferable in these Banksia woodlands. Burning rotations longer than those required for fuel reduction purposes are necessary to maximise conservation values.