Search

Worldwide, extreme climatic events such as drought and heatwaves are associated with forest mortality. However, the precise drivers of tree mortality at individual and stand levels vary considerably, with substantial gaps in knowledge across studies in biomes and continents. In 2010–2011, a drought‐associated heatwave occurred in south‐western Australia and drove sudden and rapid forest canopy collapse. Working in the Northern Jarrah (Eucalyptus marginata) Forest, we quantified the response of key overstory (E. marginata, Corymbia calophylla) and midstory (Banksia grandis, Allocasuarina fraseriana) tree species to the extreme climate event. Using transects spanning a gradient of drought impacts (minimal (50–100 m), transitional (100–150 m) and severe (30–60 m)), tree species mortality in relation to stand characteristics (stand basal area and stem density) and edaphic factors (soil depth) was determined. We show differential mortality between the two overstory species and the two midstory species corresponding to the drought‐associated heatwave. The dominant overstory species, E. marginata, had significantly higher mortality (~19%) than C. calophylla (~7%) in the severe zone. The midstory species, B. grandis, demonstrated substantially higher mortality (~59%) than A. fraseriana (~4%) in the transitional zone. Banksia grandis exhibited a substantial shift in structure in response to the drought‐associated heatwave in relation to tree size, basal area and soil depth. This study illustrates the role of climate extremes in driving ecosystem change and highlights the critical need to identify and quantify the resulting impact to help predict future forest die‐off events and to underpin forest management and conservation.

The pathogenicity of some Phytophthora species recently described from Western Australia, together with P. cinnamomi as a control, was tested against seven Western Australian native plant species in the glasshouse. Host species were Banksia grandis, B. littoralis, B. occidentalis, Casuarina obesa, Corymbia calophylla, Eucalyptus marginata and Lambertia inermis. Twenty-two Phytophthora species were grown on a vermiculite, millet seed and V8 substrate and used as soil inoculum when the plant hosts were approximately 3 months old. Pathogenicity was assessed after 6 weeks and plants were scored for death, root damage, and percentage reduction of shoot growth compared with control plants. The pathogenicity of P. cinnamomi was confirmed. Phytophthora niederhauserii was shown to be similar to P. cinnamomi in pathogenicity and of concern ecologically. Other species that killed one or more hosts were P. boodjera, P. constricta, P. elongata, P. moyootj and P. rosacearum, while P. condilina, P. gibbosa, P. gregata, P. litoralis and P. 'personii' caused significant reduction to shoot and/or root growth, but did not kill plants. Host species susceptible to the highest number of Phytophthora species were B. grandis, B. littoralis, B. occidentalis and E. marginata. No Phytophthora species tested killed C. calophylla.

Systemic treatment of stems with injections of phosphite liquid and novel soluble capsule implants of phosphite, PHOSCAP® (phosphorous, potassium, iron, manganese, zinc, boron, copper, magnesium and molybdenum) and MEDICAP MD® (nitrogen, phosphorous, potassium, iron, manganese, and zinc), were applied to Banksia grandis and Eucalyptus marginata trees to control Phytophthora cinnamomi. Four weeks after treatment application, excised branches were under-bark inoculated with P. cinnamomi. In B. grandis, phosphite implants and liquid injections significantly reduced lesion length compared to the control, and MEDICAP MD® implants; however, there was no significant difference in lesion length between trees treated with phosphite implants and liquid injections and PHOSCAP implants. In E. marginata, phosphite implants and liquid injections significantly reduced lesion length compared to the control, PHOSCAP® and MEDICAP MD® implants. In B. grandis and E. marginata, PHOSCAP® and MEDICAP MD® implants reduced the average lesion length compared to the control; however, the interactions were not significant. Results show that both liquid phosphite injections and novel phosphite implants are effective at controlling lesion extension in B. grandis and E. marginata, caused by P. cinnamomi. Further work is required to determine if nutrient application reduces Phytophthora disease through improving plant health.

Novel species of fungi described in this study include those from various countries as follows: Australia: Banksiophoma australiensis (incl. Banksiophoma gen. nov.) on Banksia coccinea, Davidiellomyces australiensis (incl. Davidiellomyces gen. nov.) on Cyperaceae, Didymocyrtis banksiae on Banksia sessilis var. cygnorum, Disculoides calophyllae on Corymbia calophylla, Harknessia banksiae on Banksia sessilis, Harknessia banksiae-repens on Banksia repens, Harknessia banksiigena on Banksia sessilis var. cygnorum, Harknessia communis on Podocarpus sp., Harknessia platyphyllae on Eucalyptus platyphylla, Myrtacremonium eucalypti (incl. Myrtacremonium gen. nov.) on Eucalyptus globulus, Myrtapenidiella balenae on Eucalyptus sp., Myrtapenidiella eucalyptigena on Eucalyptus sp., Myrtapenidiella pleurocarpae on Eucalyptus pleurocarpa, Paraconiothyrium hakeae on Hakea sp., Paraphaeosphaeria xanthorrhoeae on Xanthorrhoea sp., Parateratosphaeria stirlingiae on Stirlingia sp., Perthomyces podocarpi (incl. Perthomyces gen. nov.) on Podocarpus sp., Readeriella ellipsoidea on Eucalyptus sp., Rosellinia australiensis on Banksia grandis, Tiarosporella corymbiae on Corymbia calophylla, Verrucoconiothyrium eucalyptigenum on Eucalyptus sp., Zasmidium commune on Xanthorrhoea sp., and Zasmidium podocarpi on Podocarpus sp. Brazil: Cyathus aurantogriseocarpus on decaying wood, Perenniporia brasiliensis on decayed wood, Perenniporia paraguyanensis on decayed wood, and Pseudocercospora leandrae-fragilis on Leandra fragilis. Chile: Phialocephala cladophialophoroides on human toe nail. Costa Rica: Psathyrella striatoannulata from soil. Czech Republic: Myotisia cremea (incl. Myotisia gen. nov.) on bat droppings. Ecuador: Humidicutis dictiocephala from soil, Hygrocybe macrosiparia from soil, Hygrocybe sangayensis from soil, and Polycephalomyces onorei on stem of Etlingera sp. France: Westerdykella centenaria from soil. Hungary: Tuber magentipunctatum from soil. India: Ganoderma mizoramense on decaying wood, Hodophilus indicus from soil, Keratinophyton turgidum in soil, and Russula arunii on Pterigota alata. Italy: Rhodocybe matesina from soil. Malaysia: Apoharknessia eucalyptorum, Harknessia malayensis, Harknessia pellitae, and Peyronellaea eucalypti on Eucalyptus pellita, Lectera capsici on Capsicum annuum, and Wallrothiella gmelinae on Gmelina arborea. Morocco: Neocordana musigena on Musa sp. New Zealand: Candida rongomai-pounamu on agaric mushroom surface, Candida vespimorsuum on cup fungus surface, Cylindrocladiella vitis on Vitis vinifera, Foliocryphia eucalyptorum on Eucalyptus sp., Ramularia vacciniicola on Vaccinium sp., and Rhodotorula ngohengohe on bird feather surface. Poland: Tolypocladium fumosum on a caterpillar case of unidentified Lepidoptera. Russia: Pholiotina longistipitata among moss. Spain: Coprinopsis pseudomarcescibilis from soil, Eremiomyces innocentii from soil, Gyroporus pseudocyanescens in humus, Inocybe parvicystis in humus, and Penicillium parvofructum from soil. Unknown origin: Paraphoma rhaphiolepidis on Rhaphiolepsis indica. USA: Acidiella americana from wall of a cooling tower, Neodactylaria obpyriformis (incl. Neodactylaria gen. nov.) from human bronchoalveolar lavage, and Saksenaea loutrophoriformis from human eye. Vietnam: Phytophthora mekongensis from Citrus grandis, and Phytophthora prodigiosa from Citrus grandis. Morphological and culture characteristics along with DNA barcodes are provided.

The food resource utilisation of six species of Banksia by the endangered Carnaby’s cockatoo (Calyptorhynchus latirostris) was investigated on the Swan coastal plain, Western Australia, over a 12-month period. The energy yield from the seeds harvested by the cockatoos was determined and the information was combined with data on the number of infructescences produced per hectare, the average seed yield per infructescence and the average rate of harvest of that species of seed by the cockatoos to calculate estimates of the number of infructescences required to support a single cockatoo per day under a range of scenarios. Over 65% of infructescences of each species of Banksia handled by the cockatoos were consumed for seed. Banksia sessilis had the largest number of infructescences and follicles manipulated by Carnaby’s cockatoos. The energy content of Banksia seed was 20–23 kJ g–1. Seed weight varied from 0.075 ± 0.016 (s.e.) g for B. attenuata to 0.007 ± 0.002 (s.e.) g for B. sessilis. The number of infructescences required to meet the birds’ daily energy intake ranged from 14 for B. grandis to 3821 for B. sessilis. The results have important implications for the continued capacity of the Swan coastal plain to support Carnaby’s cockatoos, for the future survival of obligate seeding Banksia spp. and for anthropogenic revegetation programs utilising Banksia spp.

Ex-situ comparison of Phytophthora cinnamomi habitat suitability of soils from a topographic gradient across the Swan Coastal Plain from dunes to a forested peneplain provides clues to the influences of soil type on disease development. Only ex-situ testing was possible to compare P. cinnamomi habitat suitability because of confinement of Swan Coastal Plain plant communities to a relatively few, small remnants. Disease expression was determined in a shadehouse environment and stimulation of sporangium production in a controlled environment. Greater than 90 % mortality of Banksia grandis following P. cinnamomi ex-situ soil inoculation and greatest stimulation of sporangium formation and release of zoospores by soil extracts from Bassendean Dune, Pinjarra Plain, Ridge Hill Shelf and Jarrah Forest reflected high disease incidence and severity that occurs in these landscape elements. In Quindalup Dune (QD) soil there were no deaths of B. grandis following P. cinnamomi soil inoculation, consistent with the lack of disease incidence and expression that occurs in QD plant communities. However QD soils stimulated sporulation similar to the other landscape element soil extracts. Higher levels of phosphorus and calcium in QD soils possibly influenced the host-infection process after sporangium formation and zoospore release. Greater than 90 % plant mortality following P. cinnamomi soil inoculation and stimulation of sporulation by Spearwood Dune (SD) soil extracts did not reflect low disease incidence and severity that occurs in this landscape element. Differences in measured variables between soils did not indicate reasons for the high plant mortality following ex-situ soil inoculation and low disease expression in SD plant communities, possibly due to post sporulation effects of the soil environments on host infection and invasion by P. cinnamomi.

Drought and heat-induced forest dieback and mortality are emerging global concerns. Although Mediterranean-type forest (MTF) ecosystems are considered to be resilient to drought and other disturbances, we observed a sudden and unprecedented forest collapse in a MTF in Western Australia corresponding with record dry and heat conditions in 2010/2011. An aerial survey and subsequent field investigation were undertaken to examine: the incidence and severity of canopy dieback and stem mortality, associations between canopy health and stand-related factors as well as tree species response. Canopy mortality was found to be concentrated in distinct patches, representing 1.5 % of the aerial sample (1,350 ha). Within these patches, 74 % of all measured stems (>1 cm DBHOB) had dying or recently killed crowns, leading to 26 % stem mortality six months following the collapse. Patches of canopy collapse were more densely stocked with the dominant species, Eucalyptus marginata, and lacked the prominent midstorey species Banksia grandis, compared to the surrounding forest. A differential response to the disturbance was observed among co-occurring tree species, which suggests contrasting strategies for coping with extreme water stress. These results suggest that MTFs, once thought to be resilient to climate change, are susceptible to sudden and severe forest collapse when key thresholds have been reached.

The ability of Phytophthora cinnamomi to survive long dry periods is the key to its persistence in the south-west of Western Australia. It has been proposed that dead Banksia grandis are a significant long-term reservoir for P. cinnamomi inoculum. To test this, 36 healthy B. grandis trees were inoculated in April 1999, and the presence of viable propagules in planta was determined between 2 and 34 months after tree death. By 10 months after inoculation, 75% of the trees had died, with the remaining seven trees dying by 22 months. The pathogen was more commonly recovered from bark than from wood, except from those trees that died at 22 months, and more commonly from above-ground trunks than below-ground trunks and roots until 8 months after plant death. In trees that died 12 months after inoculation, P. cinnamomi was recovered from 60% of trunk and root core samples at 3 months, declining to 33% at 10 months, 5.5% at 12 months and 0.1% at 34 months after tree death. In trees that died at 22 months, P. cinnamomi was recovered from 87% of trunk and root samples 2 months after tree death, decreasing to 0.5% by 33 months. This study suggests that the pathogen does not have a saprotrophic phase within dead B. grandis tissue, and B. grandis is unlikely to be a long-term reservoir for P. cinnamomi. However, the manipulation of the density of B. grandis and the use of fire to facilitate the breakdown of dead Banksia trunks in the Eucalyptus marginata (jarrah) forest may reduce the spread and impact of P. cinnamomi.

Rates of weight loss and release of N, P, K, S, Ca, Mg, Na, and Cl from litter of several species in jarrah (E. marginata Donn ex Sm.) forest were measured in relation to site fire history and soil type. Weight loss from leaf litter decreased in the order jarrah > marri (E. calophylla R. Br. ex Lindl.) > Banksia grandis Willd. After 18 months on the forest floor senesced leaves of jarrah, marri and Banksia had lost 45%, 42% and 19%, respectively, of their original weight. Although greatest rates of decomposition occurred on a site burnt 3 y previously by an intense autumn fire and slowest rates on a site which had not been burnt for 8 y, the differences between burn sites were small in comparison with the total weight loss from decomposing litter. The order of release of nutrients from decomposing eucalypt litter was P dark sandy duplex soil > yellow gravels. These differences may be related to the higher nutrient status of the reddish gravel soils.

Summary The effect of phosphite concentration on lesion development by Phytophthora cinnamomi in stems and roots of Banksia grandis and Eucalyptus marginata and in stems of Banksia coccinea was assessed during a 4.3 year period after stem injection of phosphite. Lesion length 6 weeks after inoculation was significantly less in roots of B. grandis trees that had been stem injected with three concentrations of phosphite (50, 100 and 200 g phosphite/l) at two rates (1 and 2 ml/cm of stem circumference) compared with the not-injected control. With the exception of B. grandis trees injected with 50 g phosphite/l, lesion length for the high rate was not significantly different to the low rate. In roots of E. marginata, lesion development in response to phosphite was different to that in roots of B. grandis; lesion length in roots did not differ significantly between phosphite concentration and rate. Lesion length and girdling in stems of B. grandis and E. marginata was significantly less in those injected with phosphite than in not injected stems. One year after injection, callus tissue had contained lesions in stems injected with phosphite. By 4.3 years after injection of both hosts there was a steep significant negative linear relationship between phosphite concentration and either lesion length or girdling, with greatest lesion development in not injected stems and least in stems injected with 100 g phosphite/l. Recovery of P. cinnamomi from lesion margins 1 year after injection, was significantly less in trees injected with phosphite than in not injected trees. The amount of plant death reflected containment of lesion extension and girdling, and reduction of recovery of P. cinnamomi with phosphite concentration; 4.3 year after injection there was a steep significant negative linear relationship between phosphite concentration and percentage of plant death. In contrast to B. grandis and E. marginata, there was a U-shaped non-linear relationship between phosphite concentration and effectiveness of phosphite in controlling lesion extension and girdling in B. coccinea. Containment of lesion extension and girdling with time was greatest for B. coccinea stems injected with 25 g phosphite/l, least for stems not injected, and intermediate in stems injected with 50 and 100 g phosphite/l. As in B. grandis and E. marginata, containment of lesion extension and girdling in B. coccinea with phosphite concentration was reflected in the amount of plant death. The non-linear response to phosphite of some plant species indicated that injected concentration for B. coccinea should not exceed 50 g phosphite/l, whereas injected concentrations of up to 100 g phosphite/l could be recommended for B. grandis. Longevity of action of phosphite for 4–5 years in native plant species after one injection makes phosphite injection a practical control option for the control of P. cinnamomi disease front extension and the protection of threatened flora. Research into the effect of factors affecting longevity of action of phosphite would facilitate optimization of timing of injection. Resume L'effet de la concentration de phosphite sur le developpement des lesions de Phytophthora cinnamomi Rands dans les troncs et racines de Banksia grandis Wild. et Eucalyptus marginata Donn. ex Smith et les troncs de B. coccinea R. Br. a eteetudie sur une periode de 4.3 annees apres injection du phosphite dans les troncs. La longueur de lesion est significativement plus faible dans les racines de B. grandis ayant subi une injection 3 semaines apres inoculation avec 3 concentrations de phosphite (50, 100 et 200 g de phosphite/l) a deux taux (1 et 2 ml/cm de circonference du tronc) que chez les temoins non traites. A l'exception des B. grandis injectes avec 50 g de phosphite/l, la longueur des lesions pour le taux d'injection le plus fort est non significativement differente de celle obtenue pour le taux le plus faible. La reponse au phosphite est differente dans les racines de E. marginata par rapport a celles de B. grandis; la longueur de lesion n'est dans ce cas pas affectee significativement par la concentration de phosphite et le taux d'injection. La longueur de lesion et la ceinturation des troncs sont plus faibles chez les B. grandis et E. marginata injectes que chez les temoins. Un an apres injection, les lesions sont efficacement contenues par les tissus de cicatrisation dans les troncs traites. Apres 4.3 annees, on observe une relation lineaire significative fortement negative entre la concentration de phosphite et la longueur de lesion ou la ceinturation, avec le plus fort developpement dans les troncs non traites et le plus faible dans les troncs traites a 100 g de phosphite/l. Le reisolement de P. cinnamomia la marge des lesions un an apres injection est significativement plus faible dans les arbres traites avec le phosphite que dans les arbres non traites. Le niveau de mortalite reflete la limitation de l'extension des lesions et de la ceinturation, ainsi que le moindre reisolement de P. cinnamomi avec l'augmentation de la concentration en phosphite: on observe une relation lineaire significative fortement negative entre la concentration de phosphite et le pourcentage de mortalite, 4.3 annees apres l'injection. Contrairement aB. grandis et E. marginata, une relation non lineaire est observee chez B. coccinea entre la concentration de phosphite et l'efficacite de limitation des lesions et de la ceinturation. Celle-ci est maximale pour les tiges de B. coccinea injectees a 25 g de phosphite/l, minimale pour les tiges non traitees, et intermediaire pour les tiges traitees a 50 g et 100 g de phosphite/l. Comme chez B. grandis et E. marginata, la limitation des lesions et de la ceinturation en fonction de la concentration de phosphite se traduit en terme de niveau de mortalite. La reponse non- lineaire au phosphite suggere que, dans le cas de B. coccinea, la concentration de 50 g/l de phosphite ne devrait pas etre depassee pour les injections, alors que des concentrations allant jusqu’a 100 g/l pourrait etre recommandees dans le cas de B. grandis. La persistance de l'action du phosphite pendant 4–5 ans dans des plantes indigenes apres une seule injection montre que cette methode peut etre utilisee en pratique pour lutter contre l'extension du front de la maladie causee par P. cinnamomi et proteger la flore menacee. L’etude des facteurs affectant cette persistance pourrait permettre d‘optimiser la periode d'injection. Zusammenfassung Der Einfluss der Phosphit-Konzentration nach einer Injektion im Stamm auf die Bildung von Lasionen durch Phytophthora cinnamomi an Wurzeln und Stammen von Banksia grandis und Eucalyptus marginata sowie an Stammen von Banksia coccinea wurde wahrend 4,3 Jahren untersucht. Die Lange der Lasionen war in Wurzeln von B. grandis mit einer Phosphitbehandlung (3 Wochen nach Inokulation der Wurzeln Stamminjektion mit 50, 100 und 200 g/l Phosphit in zwei Dosen von 1 und 2 ml/cm Stammumfang) signifikant kleiner als in unbehandelten Kontrollbaumen. Mit Ausnahme von 50 g/l Phosphit wirkte sich die hohere Dosis im Vergleich zur tieferen signifikant auf die Lange der Lasionen aus. An Wurzeln von E. marginata hatten dagegen weder die Phosphitkonzentration noch die beiden Dosen einen signifikanten Effekt auf die Lasionenbildung. Nach einer Stamminokulation bewirkten alle Phosphitbehandlungen kurzere Lasionen und weniger stammumfassende Nekrosen bei B. grandis und E. marginata im Vergleich zu den unbehandelten Kontrollen. Ein Jahr nach der Injektion waren die Lasionen an Stammen mit Phosphitbehandlung durch Kallusgewebe abgegrenzt. Nach 4,3 Jahren ergab sich fur beide Arten eine enge, signifikant negative lineare Beziehung zwischen der Phosphitkonzentration und der Lange und Breite der Nekrosen, mit den grossten Lasionen an unbehandelten Stammen und den kleinsten an solchen, die mit 100 g/l Phosphit behandelt worden waren. P. cinnamomi wurde aus den Lasionsrandern ein Jahr nach der Injektion bei behandelten Baumen im Vergleich zur Kontrolle signifikant weniger haufig nachgewiesen. Die Begrenzung der Lasionen und die Reduktion von P. cinnamomi wirkte sich auch auf die Mortalitat aus: 4,3 Jahre nach der Injektion ergab sich eine enge, signifikant negative lineare Beziehung zwischen der Phosphitkonzentration und dem Prozentsatz abgestorbener Baume. Im Gegensatz zu B. grandis und E. marginata ergab sich fur B. coccinea eine U-formige nicht lineare Beziehung zwischen den Phosphitkonzentrationen und der Wirksamkeit der Behandlung. Hinsichtlich des Wachstums der Lasionen und des Ringelungseffektes war die Behandlung mit 25 g/l Phosphite am wirksamsten, 50 und 100 g/l waren intermediar und Baume ohne Injektion reagierten am wenigsten. Dies galt auch fur die Mortalitat. Aufgrund der nicht linearen Reaktion einiger Pflanzenarten auf Phosphit sollten auch bei B. coccinea 50 g/l Phosphit nicht uberschritten werden, bei B. grandis werden 100 g/l empfohlen. Die lange Wirksamkeit von 4–5 Jahren bei einheimischen Pflanzenarten nach nur einer Injektion lasst Phosphite als geeignet erscheinen, um die Ausbreitung von P. cinnamomi an einer Krankheitsfront zu bekampfen und die gefahrdete Flora zu schutzen. Untersuchungen zu Einflussfaktoren auf die Dauer der Phosphitwirkung konnten die Optimierung des Injektionszeitpunktes erleichtern.

Trees growing along windy coasts often have canopies that are greatly reduced in size by the sculpting effects of wind and salt spray. Trees with environmentally reduced stature are called elfinwood (windswept shrub-form or krummholz) and are ecologically important because they represent outposts growing at the limit of tree success. The purpose of this study was to assess if Banksia grandis elfinwood growing at Cape Leeuwin had a different nutrient status than normal low-form (LF) trees growing nearby, and if nutrient deficiencies, toxicities and/ or imbalances were among the limiting factors imposed on elfinwood. The concentrations of N, P, K, Ca, Mg, Na, Cl-, Fe, Mn, Zn, Cu, Mo and B were analysed for mature green foliage, immature foliage, foliage litter, flowers and soil. When the elfinwood and LF trees were compared, the foliar nutrient status was generally similar, except that elfinwood foliage had significantly higher mean concentrations of N, Zn and Cu, while LF trees had higher Fe and Mn contents. Many nutrients were conserved before leaves were shed in both elfinwood and LF trees, including N, P, K, Na, Cl - , Mn and Cu (LF trees also conserved Ca and Mg). However, elfinwood and LF tree-litter contained significantly higher Fe concentrations than green foliage (elfinwood litter also had higher levels of Mg and B). It is tempting to suggest that the translocation of Fe into leaves before they were shed is a regulation mechanism to prevent Fe toxicity, or imbalance in the Fe : Mn ratio. Proteoid roots strongly acidify the soil to mobilize P, which also chemically reduces Fe +3 to plant-available Fe +2 . The increased supply of Fe +2 in the rhizosphere, caused by the action of proteoid roots, might tend to defeat self-regulation of Fe uptake. It is possible that excess Fe accumulation in the plant might be regulated, in part, by exporting Fe into the leaves before they are shed. The nutrient status of B.grandis elfinwood is compared with mountain elfinwood of North America. The extreme habitat of coastal elfinwood provides many theoretical pathways for nutrient limitation, but B. grandis elfinwood at Cape Leeuwin does not appear to be nutrient deficient.

Roots of a wide range of plant species exude carboxylates, e.g. citrate, into the rhizosphere, to mobilise sparingly available phosphate. We investigated the carboxylates in root exudates of Banksia grandisWilld. (Proteaceae), which occurs on severely phosphate-impoverished soils in Western Australia. Plants were grown in pots with a nutrient-poor quartz sand, with phosphate, at 25 μg P g−1, added as either K-phosphate, glycerol phosphate, Fe-phosphate or Al-phosphate.

The efficacy of phosphite to control the production of zoospores of Phytophthora cinnamomi on infected trees grown in a glasshouse and in a revegetated mined area was examined. Banksia grandis and Eucalyptus marginata seedlings in the glasshouse and E. marginata seedlings in the minepit were sprayed with 0, 5 and 10 g phosphite L−1. In both trials, zoospores were produced from infected tissue of plants treated with all concentrations of phosphite. In the glasshouse, spray application of 5 and 10 g phosphite L−1 significantly reduced the production of zoospores from both B. grandis and E. marginata seedlings. In the mined area there was a similar, though nonsignificant, reduction in the number of zoospores produced from phosphite-treated and nontreated E. marginata seedlings. However, the average number of zoospores produced was greater in plants not treated with phosphite (1·75 zoospores mL−1) than from plants treated with 5 or 10 g phosphite L−1 (0·04 and 0·09 zoospores mL−1, respectively). Pimelea ferruginea leaves were used to bait the water surrounding the plants in the mined area to determine if zoospores produced from phosphite-treated plants were able to infect plant material. Significantly more baits were infected by zoospores from plants not treated with phosphite compared with plants treated with 5 or 10 g phosphite L−1. These results suggest that phosphite reduces, but does not prevent, the production of viable zoospores on infected trees. Thus phosphite application may not remove the risk of P. cinnamomi spreading from infested, sprayed areas.

Roots of a wide range of plant species exude carboxylates, such as citrate, into the rhizosphere. In the present study, seedlings of a range of Australian Banksia, Hakea and Dryandra species (Proteaceae) were assayed for their exudation of carboxylates. All of these species (Hakea prostrata, Hakea undulata, Hakea petiolaris, Hakea baxteri, Banksia grandis, Banksia prionotes, Banksia occidentalis and Dryandra sessilis) form cluster roots when grown in nutrient solution with a low phosphate concentration. Exudation of carboxylates was studied for cluster roots and non-cluster roots separately, and for the entire root system. Cluster roots of these Proteaceae exuded malate, malonate, lactate, acetate, maleate, citrate, fumarate, cis- and trans-aconitate. The relative contributions of each of these carboxylates differed between species. Malate, malonate, lactate, citrate and trans-aconitate, however, were invariably present in large proportions of total carboxylate exudation. Non-cluster roots of H. prostrata exuded a spectrum of carboxylates (mainly malonate, lactate and citrate), which differed somewhat from the exudation pattern of cluster roots (mainly malate, malonate, lactate and citrate). The rate of exudation for cluster roots of the seven species was approximately 1.6 nmol g -1 FM s -1 , which is considerably higher than that reported for a variety of crop and native species that do or do not form cluster roots. Contrary to what occurs in the cluster roots of Lupinus albus, which release carboxylates accompanied by protons so that the rhizosphere is acidified, the present Proteaceae exude the carboxylates as anions without concomitant proton release. The role of carboxylates in the mobilization of phosphate and other nutrients from soil is discussed.

Dry sclerophyll forest dominated by Jarrah (Eucalyptus marginata Donn ex Sm.) covers about 1.6 Mha of south-western Australia and is, relative to other eucalypt forests, low in many nutrients, especially N and P. If fire is excluded from these forests, Banksia grandis Willd. (Proteaceae) often grows in dense thickets as the dominant understorey. Jarrah has an extensive surface (0-20 cm) system of fine lateral roots with ectomycorrhizal associations, while B. grandis produces a mat of cluster (proteoid) roots (0-20 cm) in the late winter and early spring. We measured seasonal changes in acid phosphatase activity and concentrations of ergosterol, microbial P and other P fractions (NaOH-extractable inorganic and organic P, Bray inorganic P) in soil from a forest where B. grandis was either present (Jarrah + Banksia) or absent (Jarrah). Acid phosphatase activity was between 30 and 40 μmol p-NP g -1 h -1 in the moist winter and spring, but declined to less than 10 μmol p-NP g -1 h -1 during the dry summer. Microbial P varied from less than 10 μg g -1 in the late summer to more than 50 μg g -1 during the wetter times of the year. Ergosterol also increased about four-fold as soil moisture content increased (P ≤ 0.05). Acid phosphatase was significantly and positively related to both ergosterol and microbial P in both Jarrah + Banksia and Jarrah soils, but the intercepts and slopes of the regressions varied seasonally and were not significantly different (P ≤ 0.05) between species composition. About 50% of the variation in phosphatase activity in Jarrah soils was explained by variation in fungal biomass (ergosterol), regardless of season, while in Jarrah + Banksia soils, 74% was explained by fungal biomass when soils were dry, decreasing to less than 10% as soil moisture increased. Linear regression models showed that ergosterol and microbial P are good predictors of acid phosphatase activity in soil, but their relative importance is dependent on plant species composition and season. We propose that the relative importance of fungal biomass as a determinant of phosphatase activity in the soil decreases as the seasonal flush of cluster roots dominate in the Jarrah + Banksia soils, while in Jarrah soils the phosphatase flush is primarily due to increased activity of the fungal biomass. Seasonal and spatial heterogeneity in microbial P, ergosterol and in the source of acid phosphatase in the soil are therefore a consequence of plant species composition and root type and will influence nutrient availability at the ecosystem scale.

Analysis of globoid-held mineral nutrients and aluminum in Banksia grandis seeds collected from two sites, one with a high aluminum content in the soil, demonstrated that there was no significant difference in the concentration or distribution of mineral nutrients between seed stocks. Even in seeds from high-aluminum soils, aluminum was not detected in the globoids and only traces of the metal were noted within calcium-rich crystals. It was concluded that a conservative regulation of nutrients occurred in B. grandis, and even in a family known for it's aluminum-accumulating potential the metal was largely excluded from the propagules. Key words: Banksia grandis, Proteaceae, aluminum, globoids, energy-dispersive X-ray microanalysis.

Three polyvalent Streptomyces phages were used to isolate four Micromonospora species (M. carbonacea, M. chalcea, M. purpureochromogenes, and M. inositola) from mine-site rhizosphere soils in Western Australia. Streptomyces violascens was isolated using selective isolation techniques from the same soils. The Micromonspora spp. were examined for their ability to produce cellulases. Micromonospora carbonacea, M. chalcea, and M. purpureochromogenes, which were found to produce the enzyme, caused lysis of Phytophthora cinnamomi hyphae. Glasshouse trials showed that the use of the cellulase-producing M. carbonacea isolate, in conjunction with the antibiotic-producing S. violascens isolate, had a synergistic effect on the suppression of the Phytophthora root rot and in promoting growth of Banksia grandis. The importance of using a number of antagonists with different antagonistic abilities to control plant pathogenic fungi is discussed. Keywords: biological control, Micromonospora carbonacea, Streptomyces violascens, cellulases, Phytophthora cinnamomi.

Pollen protoplasts may supply important nutritional resources for birds; however, they are locked up within the mechanically strong and biochemically complex pollen wall. Previous studies of pollen digestion in birds have yielded highly variable and often contradictory results. We tested whether these differences could reflect the vastly different methodologies that have been used. We used a standard method to investigate digestion of Banksia grandis (Proteaceae) pollen in New Holland honeyeaters (Phylidonyris novaehollandiae). Four types of B. grandis pollen were examined: fresh, frozen for a week, stored in conditions to stimulate pregermination for 24 h, or collected by honeybees. Our data indicate that although pollen treatment may influence digestibility of the pollen grains, these differences do not reach statistical significance because they are dwarfed by a high degree of variability between birds fed the same diet as well as variability in gut transit time (generally more pollen grains were digested over longer transit times). Similar patterns were observed for red wattlebirds (Anthochaera carunculata) fed bee-collected pollen. We believe that feeding behaviour or gut transit time may explain the marked differences between previous studies of pollen digestion by nectarivores, particularly the conflicting results for New Holland honeyeaters.

The influence of site on effectiveness of low-volume spray of phosphite for the control of Phytophthora cinnamomi was compared between sites differing in soil nutrient status. The sites ranged from infertile deep grey sands to a red loam in the Esperance Plains bioregion, and infertile gravelly sand of the Havel site-vegetation type P to a red loam of Havel site-vegetation typeQin the northern Jarrah Forest bioregion. Following low-volume spray of 0, 24 or 48 kg phosphite/ha, phosphite effectiveness was determined from assessment of rate of colonisation and inhibition of stem colonisation by P. cinnamomi, and stem phosphite concentration in challenge inoculation stems of Lambertia inermis var. inermis in Esperance Plains bioregion sites and Banksia grandis in Jarrah Forest bioregion sites. Phosphite effectiveness was mainly influenced by plant species rather than site. Phosphite spray significantly controlled P. cinnamomi colonisation in B. grandis, but not in L. inermis var. inermis. Site had no consistent influence on the effect of phosphite on P. cinnamomi despite large differences in soil nutrient status between sites. Differences between sites and site ranking changed with inoculation time and rate of phosphite spray, and duplicate sites differed significantly from each other. Site, rate of phosphite spray, time after spray and plant species significantly affected stem phosphite concentrations. In both Esperance Plains bioregion and Jarrah Forest bioregion sites, stem phosphite concentrations were greatest in fertile loam sites and least in infertile sand or gravelly sand sites. Stem phosphite concentrations increased with rate of phosphite spray with concentrations in stems receiving 48 kg phosphite/ha being 1.2 to 3-times-greater than that in stem sprayed with 24 kg phosphite/ha. Six months after low-volume spray, phosphite concentrations in L. inermis var. inermis stems had declined for all sites and spray rates to only 4–17% of original concentrations. Decline of phosphite concentrations in B. grandis stems was slower than that for L. inermis var. inermis stems. The results indicate that effectiveness of phosphite against P. cinnamomi infection in different communities will depend more on plant species composition than soil nutrient status. Differences in phosphite effectiveness between plant species may be related to differences in phosphite concentration thresholds yet to be quantified, above which tissue concentrations must be achieved before inhibition of P. cinnamomi colonisation occurs. In rapid colonisation/phosphite low-responsive species such as L. inermis var. inermis, the threshold may be greater than that required for a response in a rapid colonisation/phosphite high-response species such as B. grandis. These results demonstrate the need to determine thresholds for individual plant species and in different environments.

Disease progress curves were evaluated for the assessment of the susceptibility of the flora of the South-West Botanical Province of Western Australia threatened by Phytophthora cinnamomi infection. Disease progress was analysed with the logistic model because this model describes numerous observed disease progress curves. In addition, the three logistic model parameters, upper asymptote (Kmax), lag time (t1/2K) and intrinsic rate of increase (r), have rational physical interpretations. Because the logistic model parameters for percentage of plants with collar lesions were significantly related to parameters for percentage mortality, only the logistic model parameters for percentage mortality were used in subsequent analysis. Susceptible hosts had the greatest Kmax, shortest t1/2K and fastest r. These parameters change to lowest Kmax, longest t1/2K and slowest r for resistant plant taxa. There was a greater change of Kmax with t1/2K than with r. The Kmax and r parameters did not differ significantly between isolates of the pathogen. Variation in mortality curve parameters between years was greater for Banksia grandis than for the more susceptible B. brownii. There was no significant linear relationship between mortality curve parameters for B. brownii and temperature variables. For B. grandis there were four significant linear relationships between a mortality curve parameter and a temperature variable. Potting mix soil favoured greatest disease progress with the shortest t1/2K and fastest r. Susceptibility to P. cinnamomi determined in a shadehouse environment following soil inoculation was significantly positively correlated with susceptibility recorded in disease centres in natural environments.

Longevity of effectiveness of stem-injected phosphite against Phytophthora cinnamomi was determined in two ways. Injected and not injected Banksia grandis and Eucalyptus marginata were challenge inoculated with P. cinnamomi at different times over an 8-year period. The mortality of injected and not injected B. attenuata and B. grandis was monitored over a 16-year period on active disease fronts. Duration of phosphite effectiveness was determined from the time after stem injection to 50% inhibition of either lesion length or girdling following challenge inoculation with P. cinnamomi and time to 50% mortality following injection of trees in disease fronts. For both B. grandis and E. marginata, time to 50% inhibition of lesion length increased from 4–5 years for 50 g phosphite/L to at least 10 years for 200 g phosphite/L. Time to 50% inhibition of girdling was significantly less than that for lesion length. In B. grandis, time to 50% inhibition of girdling increased with phosphite concentration from 2.9 years for 50 g phosphite/L to 10 years for 200 g phosphite/L. Time to 50% inhibition of girdling in E. marginata was 4 years. Predicted response surfaces of the effect of phosphite concentration and time on inhibition of lesion length and girdling showed that there was a much more gradual decline in predicted inhibition in E. marginata than the more susceptible B. grandis. An injection of 100 g phosphite/L of Banksia trees on disease fronts delayed time to 50% mortality by 3 to 4.8 years (mean 4.1±0.6 years). At least 4 years of effectiveness of phosphite against P. cinnamomi in native communities is much greater than the ≲2 years found in agriculture. The effects of phosphate competition and adaptation to infertile environments on longevity of effectiveness of phosphite are discussed. Demonstration that the effectiveness of an injection of phosphite against P. cinnamomi will last at least 4 years in native flora confirms the practical utility of using phosphite injection for the control of P. cinnamomi in threatened communities. A greater understanding than available at present of the differences in phosphite uptake, within-plant distribution, loss and effectiveness between targeted plant species can optimise current phosphite prescriptions. Additional keywords: apparent mortality rate, host defence mechanisms, longevity, persistence, phosphonate, phytotoxicity.

Five Acacia species native to Western Australia were assessed for their potential to protect the highly susceptible species Banksia grandis Wield from infection by the plant pathogen Phytophthora cinnamomi Rands. In a rehabilitated bauxite pit at Jarrahdale 55 km south-east of Perth and in a glasshouse trial, B. grandis planted either alone or with A. pulchella R.Br., A. urophylla Benth., A. extensa Lindl., A. lateriticola Maslin or A. drummondii Lindl., was soil inoculated with P. cinnamomi. It could only be shown that A. pulchella significantly protected B. grandis from P. cinnamomi infection in the rehabilitated bauxite pit trial up to 1 year after inoculation. This confirms the potential of this species for biological control of the pathogen in infested plant communities. The observed protection was not the result of a decrease in soil temperature or moisture. Protection was not emulated in a glasshouse trial where optimum environmental conditions favoured P. cinnamomi. Despite a delay in infection of B. grandis planted with Acacia spp., none of the five species definitively protected B. grandis from P. cinnamomi. However, in the glasshouse trial, A. pulchella, A. extensa, A. lateriticola and A. drummondii did significantly reduce the soil inoculum of P. cinnamomi, indicating a possible biological control effect on the pathogen. The mechanisms of biological control are discussed and the implications for management of rehabilitated bauxite mined areas and forests severely affected by P. cinnamomi are considered.

The vegetation of seven sites in the northern jarrah forest of Western Australia infested with Phytophthora cinnamomi was recorded and compared with adjoining vegetation. The number of species per quadrat was found to be the same in vegetation affected by P. cinnamomi as in healthy vegetation, although there were more species overall in affected vegetation. Vegetation of uninfested sites had a higher cover and more species per quadrat of trees and shrubs and lower cover and fewer species per quadrat of annual plants than vegetation of infested sites. Although many species that are known to be highly susceptible to infection by P. cinnamomi were rare at infested sites, only two (Banksia grandis and Tetratheca hirsuta) were absent from all of the 50-year-old infested parts of sites. Several species that are known to be highly susceptible to infection by P. cinnamomi were as common at infested as at healthy sites. The presence of such species at infested sites and the capacity of P. cinnamomi to infect species it does not kill suggest that this pathogen will persist and continue to influence future vegetation in the jarrah forest.

The spatial distribution of Phytophthora cinnamomi Rands at seven dieback sites in the jarrah (Eucalyptus marginata Donn. ex Smith) forest of Western Australia was determined by the following two baiting techniques: in situ baiting with live Banksia grandis Willd. seedlings and ex situ baiting of sampled soil and root material. Four areas within each site were sampled, reflecting dieback age and position in the landscape. Approximate dieback ages of 50, 20 and 5 years were determined by aerial photography. The 50-year-old age class was divided into wet valley floor and dry gravelly slope. Phytophthora cinnamomi was recovered most frequently from the 5-year-old (dieback fronts) and wet 50-year-old areas by both baiting techniques. It was recovered from more than twice as many areas and about five times as many samples when in situ B. grandis baits were used compared with ex situ soil and root baiting. Almost all recoveries from in situ baits were made between October and December. From both methods, it appears that P. cinnamomi has a patchy distribution within dieback sites in the northern jarrah forest. It is easily detected only on dieback fronts and wet valley floors. On dry gravelly sites affected 20 years or more ago, P. cinnamomi is rare and may even be absent at some sites. This makes confident detection of the pathogen difficult. In situ baiting at least allows a temporal component to the sampling and will be a useful method of detection in areas where P. cinnamomi is rare or transient.

Germination responses to multiple conditions related to fire, temperature and light were examined and interpreted in relation to the environment of the jarrah (Eucalyptus marginata Donn ex Sm.) forest and the mediterranean-type climate of south-western Western Australia. Eight of the 16 selected representatives required a boiling pre-treatment as a simulation of fire before subsequent germination conditions were imposed. Trial conditions included a range of constant incubation temperatures and either a constant dark or 12h light:12h dark illumination cycle. Species common to the understorey tended to have narrow optima to temperature (13-18°C) with the greatest germination being achieved in temperatures typical of winter. Eucalyptus calophylla, one of the canopy species, germinated over a wider range of temperatures than the understorey species tested. Several species (e.g. Acacia pulchella var. glaberrima, Calothamnus rupestris, Eucalyptus marginata, Trymalium ledifolium, Xanthorrhoea gracilis and X. preissii) were not affected by the different light conditions at lower incubation temperatures, but more seeds remained dormant under higher temperatures when exposed to light. Most species, however, germinated best under continuously dark conditions, which related to being buried under soil. Two species, Banksia grandis and Hakea amplexicaulis, showed much higher germination under light exposure conditions compared to continuously dark conditions. These two serotinous species are large seeded and their seed ecology favours large gap and soil surface establishment. The multiple requirements of particular combinations of fire, temperature and light are related to the ecology of these native Western Australian species.

The amount of litter and its nutrient composition have been measured at seven sites on various lateritic soils within the jarrah (Eucalyptus marginata Donn ex Sm.) forest near Dwellingup, Western Australia. The weight of litter accumulated during 6 years ranges from 9 tonnes/ha for forest growing on yellow sand to 18 tonnes/ha for forest on reddish gravels. The litter on the reddish gravels contains more than twice the amounts of N, P, K and S in litter on yellow sand and grey and yellow gravels. The proportion of fine material in the forest floor litter increases with total litter weight. Phosphorus, which is less mobile than other nutrients tends to accumulate in this fine component. There are large differences between the foliar nutrient levels of jarrah and Banksia grandis Willd. (e.g. P: 0.041%, 0.025%; K: 0.57%, 0.34%; Mg: 0.43%, 0.21%; Mn: 177 μg/g. 730 μg/g). However, these differences are not reflected in the litter from sites with and without B. grandis understorey. Soil differences and the predominant contribution of the overstorey to the litter appear to be the main factors affecting the litter composition.

Annual evaporation from groundflora, litter and soil of the jarrah forest was estimated from measurements of daily evaporation by ventilated chambers on several days over two separate 12-month periods. In the first year, when sampling ranged over 0.1 ha of forest, annual evaporation during daylight hours was estimated as 410 mm (0.32 rainfall). In the second year, sampling was more frequent, on a larger scale, and included the night hours. Annual evaporation was estimated at 360 mm (0.36 rainfall). Similarly, in the second year, annual evaporation from two trees of the dominant middle storey species, Banksia grandis , was estimated at 7500 and 18,9001 respectively. The leaf area of these two trees was 9.6 and 22.4 m 2 , respectively, so that annual evaporation, when expressed as mm 3 per mm 2 leaf area, was similar for both trees (mean = 820 ± 30 mm). Applying that value to all Banksia trees in a hectare of forest, and using a measured estimate of leaf area index of 0.19, the estimated annual evaporation from the Banksia component was 155 mm (0.16 rainfall). For the upland part of the forest sampled, the combined annual evaporation from the lower and middle storeys accounted for about half (0.51) of the annual rainfall. We conclude that reduced evaporation from the upper storey following clearing or thinning may be strongly counteracted by increased evaporation from the understorey due to increased availability of energy and water.

This paper reports an experimental and comparative field study of germination, early survival, and height growth of six tree species in the northern jarrah forest of Western Australia. The species are Eucalyptus marginata and Eucalyptus calophylla (comprising the overstorey) and Banksia grandis, Allocasuarina fraseriana, Persoonia longifolia and Persoonia elliptica (all understorey). Seed weight does not correlate with germination success, survival or height growth. Germination success is greatest for B. grandis (90%) and zero for Persoonia spp. Survival of caged germinants is highest for B. grandis and E. calophylla; that of uncaged seedlings is least for E. marginata, E. calophylla and B. grandis because of browsing. Over 40% of seeds of E. marginata, E. calophylla and B. grandis do not germinate because they are harvested and presumably eaten. Viability of seeds of all species exceeds 70%. Germination is influenced by type and depth of litter but not by artificial shading or root competition from forest trees. Browsing by herbivores (kangaroos) is an important cause of death of seedlings of E. marginata, E. calophylla and B. grandis. Litter type and depth affect survival of E. calophylla, B. grandis and A. fraseriana but root competition and shading are unimportant. Height growth of E. marginata, E. calophylla and B. grandis is reduced by root competition and browsers further reduce height growth of B. grandis. Litter affects growth of E. marginata and B. grandis and shading influences growth of E. marginata and E. calophylla. Poor germination explains the low abundance of both Persoonia spp. in the jarrah forest but differential seedling establishment does not explain the adult representation of the other four species.

Removal from the forest floor of seed or fruit of the six commonest species of trees in upland areas of the northern jarrah forest of Western Australia was studied experimentally. The smallest seeds (Eucalyptus marginata, Allocasuarina fraseriana) were taken by insects (presumably ants). Heavier seeds or fruits (E. calophylla, Banksia grandis, Persoonia longifolia, P. elliptica) were taken more often by vertebrates (presumably srrmll mammals). Seeds of E. calophylla and B. grandis were most preferred by vertebrates, possibly because they have the highest relative concentrations of nitrogen and phosphorus.

Summary The role of fire in reducing the abundance of Banksia grandis, a forest understorey species which is susceptible to the dieback disease Phytophthora cinnamomi, was investigated by lighting small experimental fires in jarrah forests (Eucalyptus marginata) over the dry summer months of 1980. The effect of a range of fire intensities on mortality and sprouting ability of B. grandis was examined two years after burning. The number and size of stems killed to ground level by fire was directly related to fire intensity and varied from little effect of low intensity fires (

Summary The reproductive characteristics of Banksia grandis Willd. were studied in the northern jarrah forest of Western Australia. Most flowering occurs in November, with maturation of fertile (seed-bearing) infructescences taking about 12 to 14 months. Number of inflorescences produced by a tree is not closely related to its diameter at breast height (DBH). Trees become reproductive at mean DBH of 4–5 cm and fruit successfully at mean DBH of 6.7 cm. Trees of DBH < 10 cm produce fewer fertile infructescences than trees of larger DBH. Moderate-intensity fire markedly reduces production of inflorescences for at least three flowering seasons. Less than 2% of flowers on an average inflorescence set fruit, and only 1 to 5% of inflorescences develop into fertile infructescences. The cause of the latter was investigated experimentally. Although insect larvae (Lepidoptera, Coleoptera) damage infructescences, inflorescences sprayed with insecticide do not develop differently from those not so treated. Insect damage is least in inflorescences produced in late November and early December but there is no corresponding increase in development of fertile infructescences. Ants do not protect inflorescences from insect damage and insects are of little significance as predators of seed within the follicle. Whether fruit development is limited by pollination could not be adequately assessed. Low fruit set in B. grandis may result from shortage of resources, with many flowers and inflorescences allowing B. grandis the option of setting large numbers of fruits should favourable conditions prevail.

Population fluctuations of Phytophthora cinnamomi Rands were determined over 4 years from direct plating of surface soil from eight infected Eucalyptus marginata Donn ex Smith sites and one irrigated site. In a lowland area of one site and in upland areas of seven sites, fungal recovery and population densities were highest in winter when rainfall was greatest, soil matric potential was greater than −10 kPa, and temperatures were at their lowest. Recoveries and population densities decreased as rainfall decreased. The fungus was recovered throughout the year from the lowland area where the effect of season on population fluctuations was less than at upland areas. In dry summer months the fungus could not be recovered from surface soil in upland areas. Infected Banksia grandis Willd. acted as inoculum foci when surface soil was moist. In some sites P. cinnamomi could be recovered from up to 94% of the soil samples taken near the collars of dead B. grandis . Within and between infected sites there were considerable differences between trees and sampling times in inoculum density and frequency of recovery of the fungus from soil at the collars of dead B. grandis . Recovery of P. cinnamomi 75-cm-downslope from collars of dead B. grandis was less than but significantly correlated with recoveries of the fungus from soil around the collar. Rain or irrigation in summer allowed formation of inoculum foci in surface soil which survived through winter until the next summer. Inoculum could be recovered more frequently from soil sampled between 10 and 80 cm below the soil surface than from the top 10 cm of soil.

SUMMARY Vesicular-arbuscular (VA) mycorrhizae occurred in seedlings of Eucalyptus marginata, E. diversicolor and Acacia pulchella grown in soil inoculated with Glomus fasciculatus. Infection levels were 30 to 35 % of the root length of the eucalypts and 50 to 70 % of the Acacia. When A. pulchella was grown in the same pots with the two eucalypts, its infection was not increased. Banksia grandis did not form VA mycorrhizae when grown alone or in association with A. pulchella. These results are discussed in relation to the evolution of plants in Australia and the ecological significance of VA mycorrhizae in the eucalypt forest ecosystem.

The ability of phosphite to control Phytophthora cinnamomi Rands in five Western Australian native plant species was examined. Foliar application of phosphite slowed, but did not completely inhibit, colonisation of stems by P. cinnamomi. For example, in Banksia hookeriana Meisn. inoculated 2 weeks after phosphite application, 5 g phosphite/L inhibited the growth rate of P. cinnamomi by 57% compared with the non-phosphite-treated plants. The longevity of phosphite efficacy varied with plant species. Foliar application of 5 and 10 g phosphite/L decreased the growth rate of P. cinnamomi in Dryandra sessilis (Knight) Domin. for at least 12 months after it was applied. Application rates of 5 and 10 g phosphite/L for Banksia grandis Willd. and 10 g/L for B. hookeriana were effective for at least 18 months after application. In Hibbertia commutata Steud. and Dampiera linearis R.Br., phosphite was effective for less than 6 and 12 months, respectively. In a second trial, plants were inoculated with P. cinnamomi at different time periods after phosphite was applied and time to death was recorded. There was a range of responses depending on the plant species and time of year they were inoculated. The initial levels of phosphite in roots and stems of B. grandis, B. hookeriana and D. sessilis and the rate of decrease of phosphite in these tissues differed between plant species. In general, concentrations of phosphite in stems were higher or equivalent to those in roots. This study indicates that the long-term efficacy of phosphite depends on both the plant species treated and the time of year the plants are infected with P. cinnamomi.

Seedlings of five native forest species and of Pinus radiata D. Don were grown in pots in a forest soil at a range of levels of added phosphate for up to 20 weeks. At intervals of 3 weeks, roots were examined for root hairs and mycorrhizas and the weights of plant parts and their phosphorus contents were measured. Components of growth and response were calculated from three-dimensional surfaces which had level of phosphate and time as the independent variables. Except for Banksia grandis Willd., the species responded to phosphate but the time at which the response appeared was affected by seed reserves of phosphorus. Thus response appeared much earlier for the small-seeded Eucalyptus divevsicolor F. Muell. and Acacia pulchella R. Br. than for the large-seeded Eucalyptus calophylla R. Br. and Eucalyptus marginata Donn. ex Sm. The absence of a response by the banksia seemed to arise because of a large seed reserve of phosphorus, a low relative growth rate and a very low concentration of phosphorus in the leaves for maximum photosynthesis. The rates of uptake of phosphorus per unit weight of roots were generally low but increased at a time coincident with the development of mycorrhizas. Relative growth rates and the net assimilation rates were also low and reached maximum values at low concentrations of phosphorus in the leaf. The outcome was that the level of applied phosphate needed for good early growth did not differ greatly from that of Trifolium subterraneum L.

Six disinfectants and a distilled water control were tested for their ability to suppress the growth of P. cinnamomi on Banksia grandis (Willde) leaf discs. Two immersion times (30 seconds and 120 seconds) and various concentrations of the chemicals were tested. it was found that of the six disinfectants, those based on quaternary ammonium compounds were more effective than the phenolic and sodium hypochlorlte compounds. There was no significant dlfference between the different quaternary ammonium compound based chemicals and no significant difference between the two times tested for these chemicals.

Soil dilution plate techniques were used to compare the numbers of bacteria, actinomycetes and fungi in the rhizospheres of Acacia pulchella, Banksia grandis and Eucalyptus marginata (jarrah). The most frequently isolated microorganisms and those detected in significantly different numbers in the rhizospheres of the three species were tested for their effects on sporangium production, zoospore discharge, zoospore germination and mycelial growth of the jarrah dieback pathogen Phytophthora cinnamomi. The total population of fungi in rhizosphere soil from B. grandis was much greater than that found in the rhizospheres of the other two species while the convesse was true for bacteria and actinomycetes, of which the largest populations were associated with A. pulchella. Penicillium spinulosum outnum- bered the combined population of other fungi in the Banksia rhizosphere but formed a much smaller proportion of the jarrah and Acacia rhizosphere microfloras, particularly the latter. P. spinulosum had no effect on mycelial growth or zoospore discharge in P. cinnamomi; it had some ability to stimulete sporangium production and, although it partly suppressed spore germination, the inhibitory effect was less pronounced than that noted for most other microorganisms. In contrast, microorganisms which strongly inhibited mycelial growth, zoospore discharge and germination represented a greater proportion of the Acacia rhizosphere microflora compared with the other microfloras, especially that of B. grandis. While some actinornycetes and fungi produced antibiotics that inhibited vegetative growth of P. cinnamomi in dual cultures, mycelial inhibition was often attributable to nutrient depletion of agar media by the test microorganisms. Similarly, nutrient deprivation resulting from microbial competition for substrates was also considered to be the stimulus for sporangium production in liquid media. The results are discussed in relation to previously reported suppression of P. cinnamomi in forest soils beneath stands of A. pulchella and the associated implications of this for biological control of jarrah dieback.

We inoculated excised roots under controlled laboratory conditions and inoculated stems in the field to compare the behaviour of Phytophthora cactorum, P. cambivora, P. cinnamomi A2, P. citricola, P. cryptogea A1 and A2, P. megasperma var. sojae and P. nicotianae var. parasitica in the secondary phloem of Banksia grandis and Eucalyptus marginata. Most of the Phytophthora species grew in excised roots of E. marginata at a similar rate. Of the Phytophthora species with similar rates of growth in E. marginata roots, P. cinnamomi was the only species that consistently grew faster in excised roots of B. grandis than in roots of E. marginata. The growth of the Phytophthora species in excised roots under controlled conditions was significantly correlated with growth in intact stems in the field. Over a range of temperatures between 10 and 25°C, the slope of the temperature-growth response curve for P. cinnamomi in excised roots of B. grandis was greater than that for P. citricola. At temperatures between 27 and 31°C, growth rates of P. cinnarnomi in excised roots of B. grandis were 1 cm or more per day compared with 0.3 cm per day for P. citricola. Differences in growth rate in the roots of the widespread understorey species B. grandis can be important to the epidemiology of a Phytophthora species in the E. marginata forest. Phytophthora cinnamomi with fast rates of growth in roots of B. grandis is more likely to have inoculum in the vicinity of major roots of E. marginata than are Phytophthora species with slow rates of growth.

Production of sporangia and zoospore discharge by Phytophthora cinnamomi were investigated in jarrah forest soils collected beneath Acacia pulchella, Banksia grandis and Eucalyptus marginata and incubated under conditions which excluded a direct effect of these species on the soil physical environment. Sporangium production was always significantly greater in soils from B. grandis than in soils from A. pulchella; soils from E. marginata gave intermediate results. There was also evidence that discharge of zoospores was suppressed in soils from A. pulchella. Although three isolates of P. cinnamomi differed in their abilities to sporulate per se, they responded similarly to different treatments.

In soils under forest conditions and on seedlings grown under glasshouse conditions jarrah mycorrhizae exhibited strong habitat preferences. Litter on the forest floor was the main site for the formation of the white and brown ectomycorrhizae, and removal of the litter layer by prescribed burning reduced by 90% the numbers of these two ectomycorrhizae. Black ectornycorrhizae were dominant in the mineral horizons of soil and were unaffected by the type of surface litter. In glasshouse experiments litter from different plant species significantly influenced formation of ectomycorrhizae by jarrah seedlings. Litter from Acacia pulchella R. Br. was favourable for formation of ectomycorrhizae but ectornycorrhizae were absent from Banksia grandis Willd. litter. Black ectomycorrhizae, Hysterangium- initiated ectomycorrhizae and vesicular arbuscular mycorrhizae formed on jarrah seedlings in the soils used in glasshouse studies. Addition of nitrogen and phosphorus to these soils influenced the relative abundance of mycorrhizae. It is postulated that inoculum levels, nutrient status of litter and soil and the presence of compounds inhibitory to fungal growth are critical in determining the abundance of mycorrhizae. Factors affecting formation of mycorrhizae by jarrah and the ecological implications of these results are discussed.

A study of foraging by 10 bird species suggests that selective logging of large Eucalyptus marginata will only have affected Melithreptus lunatus, but that proposed silvicultural treatments, including removal of Banksia grandis, may affect several other bird species.

Banksia grandis is a major understorey component of the jarrah forest and is a refuge of the soil fungus Phytophthora cinnamomi which kills many native plant species and threatens timber production. This paper presents data on the rate of growth of the lignotuber, shoot and stem of B. grandis. Length of the seedling lignotuber increased 0.23 cm year-1 over 3 years, that of saplings and trees increased 0.44 cm yea-1. The period for a seedling to attain 1.3 m height was calculated as c. 15 years. Short-term diameter increment of trees averaged 0.19-0.27 cm year-1, indicating that diameter at breast height of 20 cm should be attained in c. 80-100 years; this estimate is also supported by direct and indirect long- term data. Watering, shading and spraying with insecticide had no significant influence on height growth of saplings. Four years after fire, sapling coppice from large lignotubers nearly attained the height of the plant before the fire. Trees, but not saplings, grew faster after logging of forest. The significance of these growth rates is discussed in relation to two methods (mechanical removal, moderate-intensity fire) proposed to reduce the abundance of B. grandis in the jarrah forest.

Seven copper salts and 10 isolates from both mating types of Phytophthora cinnamomi have been tested, in vitro, for copper cation toxicity to mycelial mat growth and to the germination of chlamydospores. It has been shown that while the copper ion is toxic, extended periods of exposure (9-23 days) to the cation are often required to change the fungistatic activity to that of a true fungicide. This observation applies to both mycelial growth and to chlamydospore germination of P. cinnamomi. Soluble copper (CuS02.5H20) gave short-term protection of both Eucalyptus marginata and to a lesser degree Banksia grandis against infection by P. cinnamomi but insoluble copper (CuO) gave none. The rhizospheres of seedlings of E. marginata and B. grandis grown in pots contained populations of soil actinomycetes different from each other and the parent soil population. These actinomycetes showed substantial antifungal activity against Phytophthora cinnamomi in vitro. Moreover, the actinomycete populations were modified by the presence of copper and by the infection of the roots by P. cinnamomi. It is suggested that both the actinomycetes and the copper contributed to the protection against invasion by P. cinnamomi.

The resistance of 21 Eucalyptus spp. to Phytophthora cinnamomi Rands was assessed using wound inoculation methods. Aggressive lesions were observed in Eucalyptus marginata Uarrah) only. Lesion development was initially compared in E. marginata and E. calophylla (moderately resistant). Further comparisons were made of lesions in roots and stems of E. marginata, E. calophylla, E. patens and E. Wandoo growing in close proximity, in forest south-east of Perth. E. wandoo was the most resistant; the fungus failed to establish. As stems proved convenient for inoculation, summer stem inoculation trials were made to rate the resistance of 21 Eucalyptus spp. grown on a rehabilitated mine site. Lesions formed in the species of the Monocalyptus and Corymbia subgeneric groups but did not develop in species of the subgenus Symphyomyrtus. An alternative laboratory method of rating susceptibility of roots to invasion was also tested. Excised roots of three eucalypts, Banksia grandis and Xanthorrhoea preissii were inoculated and incubated at 25°C. Length of the root invaded was recorded; results were not well correlated with those from field studies.

Armillaria luteobubalina is a widespread primary pathogen in the Eucalyptus marginata forest of south-western Australia. Over 200 infection centres were identified during the 5-year period between 1981 and 1985. The fungus sporulated during June and July, usually from roots but sometimes from stems (e.g. E. calophylla). Armillaria luteobubalina basidiomes were found originating from roots of 34 plant species, with greatest incidence on roots of E. marginata. Root systems were excavated and patterns of A . Luteobubalina invasion recorded. Rhizomorphs were not found and fungal spread between hosts was via root to root contacts, Variation in host species' susceptibility to the fungus was reflected in different patterns of xylem compartmentalisation and variable amounts of cambial damage. The degree of resistance expressed at the collar or lower stem determined the fate of individuals of the various species. Lack of resistance in Eucalyptus wandoo to tangential spread of A. luteobubalina often resulted in death by the time columns of decay had advanced into the lower stem or butt. Banksia grandis, E. calophylla, E. gomphocephala, and E. marginata resisted to varying degrees. Inverted V-shaped lesions, often mis- taken for fire scars, were evidence of the ability of E. gomphocephala and E. marginata individuals to resist tangential spread and prevent girdling of stems. In stems of E. calophylla, lesions did not have a definite V shape, decay penetrated deeper and the fungus persisted longer than in those of E. marginata. Host mortality following infection was greater in the intermediate- and low-rainfall zones of the eastern E. marginata forest than in the high-rainfall zone to the west.

Native forest at Durras in south-eastern New South Wales and Jarrahdale in south-western Western Australia was examined for the presence of Phytophthora cinnamomi by two sampling and isolation techniques. With the lupin seeding baiting technique, randomly selected samples of soil and fine roots collected from the New South Wales site yielded P. cinnamomi when baited, while similar baiting of comparable samples from Western Australia failed. Direct plating of samples of upper roots and root collars of recently dead Banksia grandis from Western Australian sites yielded P. cinnamomi, while this organism was not isolated from comparable samples of chlorotic Macrozamia communis collected at the New South Wales site. The results suggest that the form of occurrence of P. cinnamomi and its association with disease in Australia vary in different situations. Viewing each situation independently might ensure the adoption of control/prevention strategies appropriate to all.