Search

We tested the hypothesis that roads in arid central Australia modifyroadside hydrology and nutrient patterns creating better habitat formistletoes (Loranthaceae). In two areas of mulga woodland near AliceSprings where mulgas (Acacia aneura) are the principal mistletoe host, we compared mistletoe abundance, and host foliar water and nutrient content between roadside and interior (100 m from the road) sites. Mistletoes were significantly more abundant close to roadsides than at interior sites and mulgas had significantly higher foliar water contents at roadside compared to interior sites. Mulga foliar P and N levels were not, however, significantly different between roadside and interior sites. We conclude that roads exert a strong influence on the adjacent biota in arid central Australia primarily through facilitating water infiltration at roadside sites. [ABSTRACT FROM AUTHOR]

Plant community composition and its likely environmental controls were investigated for 200 sample plots (each 100 m2) from Mediterranean-type vegetation in the Little Desert National Park, Victoria. TWINSPAN classification revealed four readily identifiable vegetation types; mallee-broombush, heathland, stringybark open woodland, and an assemblage intermediate between mallee-broombush and heathlandreferred to here as broom-heath. Mallee-broombush was found on Parilla Sands characterized by high Ca levels relative to heathland and stringybark open woodlands on unconsolidated Lowan Sands. The first axis of a 2 dimensional non-metric MDS ordination also divided heathlands (high axis scores) from mallee-broombush (low scores), while the second separated these vegetation types from stringybark woodlands and broom-heath. Vector-fitting revealed significant correlations betweenthe locations of samples in ordination space and exchangeable soil Ca, soil colour, aspect and Shannon-Weiner diversity. Highest species richness/diversity was associated with the ecotonal area between Parilla and Lowan Sands (i.e. broom-heath) where a number of species characteristic of different assemblages had overlapping ranges. The fire-sensitive conifer, Callitris rhomboidea, was preferentially located in stringybark woodland and broom-heath vegetation types. Its presencewas positively associated with high species richness and aspects having a southerly component. Four Callitris stands sampled for population structure were all > 40 years old and showed evidence of interfirerecruitment from seeds released by old, serotinous cones. Overall, results suggest that variations in plant community composition and structure in the eastern block of the Little Desert are primarily due tovariations in soil properties associated with the distribution of the two dominant substrate types, Parilla Sand and Lowan Sand. However,the interplay of topography and fire behaviour has probably been more impo [ABSTRACT FROM AUTHOR]

The influence of factors associated with fire on seed germination ofAustralian native species is generally well documented, but examplesinvolving the use of smoke as a fire analogue for ecological research remain limited. The role of season of treatment in the efficacy of smoke as a promotive germination agent was investigated over two growing seasons using natural soil stored seedbanks in Banksia woodland near Perth, Western Australia. Smoke was applied to unburnt sites in the autumn, winter and spring of 1994. Germinant emergence and seedling survival of 37 species representing 18 families was monitored in both unburnt sites and in adjacent, recently burnt sites until the second spring after treatment (October 1995). Recruitment from seed was found to be profoundly affected by the season in which dormancy breaking treatment had been applied. The promotive effect extended beyond the initial year of application. For the majority of the species investigated, application of smoke to unburnt sites in autumn promoted a significantly greater germination response than treatment in winter orspring. In only three cases (introduced annuals, the Fabaceae and Hibbertia amplexicaulis) did autumn smoke treatment not yield better germination than in summer-burnt counterparts. However, in almost half of the cases examined, proportions of seedlings surviving past their first summer after emergence in burnt areas were consistently greaterthan those in smoked or untreated sites. Most notably, no seedlings emerging during the spring of the first year of study survived into the following summer. Implications of the results with respect to future seed bank research and management of native vegetation are discussed. [ABSTRACT FROM AUTHOR]

Live and dead trees were measured in macropod-grazed and cattle-grazed Eucalyptus woodland in north Queensland following a severe droughtbetween 1992 and 1994. ANCOVA revealed no effect of grazing treatment on the proportion of drought death. Twenty-seven per cent of all tree species were killed by the drought and the value ranged from about4% for Corymbia dallachiana (Benth.), K.D.Hill & L.A.S. Johnson to 29% for the dominant species Eucalyptus xanthoclada Brooker and A.R.Bean, although differences were non-significant. There was also no significant difference in mortality between poles (< 10 cm d.b.h.) and trees (> 10 cm d.b.h.). The study highlights a natural (as in pre-European), catastrophic structural collapse of tropical eucalypt woodland.Presumably the dramatic declines in basal area as a result of drought are recovered during successive relatively wet years by thickening events, although clearly this severe drought and its after-effects warrant further study. [ABSTRACT FROM AUTHOR]

Grasslands encompass a broad array of vegetation and climatic zones.We describe the first developments towards a rule-based functional model for predicting vegetation structure in Australian and New Zealand pastures and rangelands. The approach aims to predict the combined effects of climate and disturbance by humans and grazing livestock, and to provide a level of resolution needed for predicting changes in pastures and rangelands. We enlisted expert knowledge to develop: (l)a minimum set of critical traits; (2) rules relating site variables to favoured-plant attributes; (3) rules relating attributes to plant functional traits, and (4) rules relating plant functional types to likely plant communities. We tested the resulting model by deriving some simple predictions of plant communities of some existing pasture and rangeland sites in Australia and New Zealand, with differing climatic and human disturbance inputs. The results indicate that this first model is able to predict plant communities with varying success rates, and with the best results in cases where there are extreme climates or high management inputs. Key sensitivities in the model where further research is required include: (1) the urgent need for more explicit understanding of the key plant functional attributes favoured bydiffering climates and disturbance regimes, (2) the functional relationships between these plant functional attributes and recognisable plant functional types in vegetation, and (3) the assembly rules for the coexistence of these different plant functional types in major plant communities. The same understanding is required for subsequent process-based modelling development. [ABSTRACT FROM AUTHOR]

This paper summarizes trends in Australian wetland rehabilitation onthe basis of responses to a questionnaire from sixty-nine rehabilitation projects and a literature review. Project sizes ranged from 0.4 to 110,000 ha. Costs ranged from A $6 to 70,000 ha-1 yr-1, with most projects costing less than A$1,000 ha-1 yr-1 and with larger projects generally costing less per area than smaller projects. The oldest project began in 1963, but most projects began between 1990 and 1996. The most commonly cited dominant plant genera in rehabilitation projects were Eucalyptus, Melaleuca, Muehlenbeckia, Juncus, Avicennia, and Typha. Filling and draining and altered hydrology were the most commonly cited impacts that led to a need for rehabilitation, while excavation and planting were the most commonly cited rehabilitation methods. About 65% of projects undertake rehabilitation methods that address the causes of impacts. Habitat improvement was the most commonly stated objective in wetland rehabilitation. Of the projects that supported monitoring programs, fewer than 25% appeared to monitor variables that were closely linked to project goals. Thirty-six projects reported involvement in research other than monitoring, with vegetation research most frequently cited. Most projects encouraged some level of community involvement. Wetland rehabilitation in Australia could benefit from a centralized register of projects and a library of rehabilitation literature, including government and in-house reports. [ABSTRACT FROM AUTHOR]