Search

An understanding of rarity and how it relates to extinction risk is a central concern of conservation biology. Classic conceptions of rarity revolve around spatial rarity, based on distribution and abundance, rather than temporal rarity, where species may be common following certain conditions but rare for most of the time. This form of rarity is likely to be especially prevalent in highly variable arid systems. Rarity in the arid zone is also characterised by poorly understood threats, such as grazing, and may also reflect low collection effort given the vast and inaccessible areas involved. This study explores rarity and threat in the arid zone, based on the flora of a large region of western Queensland. The status of all species known to occur in the study area was systematically assessed, and the current list of threatened species was examined for bias in forms of rarity, life forms and habitats. Five threat syndromes were identified, arising from the interaction of plant biology and threatening processes. Over 60 potentially threatened species have been overlooked in the listing process. The list is dominated by narrow endemics from residual and spring habitats and the species from springs at least are genuinely threatened. Widespread but sparsely occurring species are under-represented in the current list, as are grasses. With the exception of spring-dependent species, plant conservation in western Queensland is currently constrained by lack of basic data on distribution, abundance, population dynamics and realistic threat syndromes for nearly all species. Separating the influence of genuine rarity, temporal rarity and low collection effort, as well as a more detailed understanding of threatening processes are needed to address plant conservation in the arid zone.

There is a prevailing paradigm that woody vegetation is expanding at the expense of grassland with reduced burning under pastoralism in the Mulga Lands biogeographic region in eastern Australia. This raises the possibility that the region is acting as a carbon sink. Vegetation boundaries were precisely positioned from rail survey plans dating from 1895 to 1900. This baseline was compared with the position of boundaries on 1952 aerial photography and 2010 Google Earth imagery. The conversion of forest to non-forest by mechanical clearing was also mapped from satellite imagery. There was no consistent trend in the direction of boundary movement for mulga (Acacia aneura F.Muell. ex Benth.), gidgee (Acacia cambagei R.T. Baker) forest or miscellaneous other forest types. The stability of the boundaries, despite the transition from aboriginal management to rangeland pastoralism, contrasts with dramatic declines in tree cover resulting from mechanical clearing. Mapping of forest cover from satellite imagery reveals that conversion of forest to non-forest has reduced mulga forest to 74%, gidgee forest to 30% and miscellaneous forest types to 82% of their original area. Annual clearing rates for the period between 1997 and 2005 were 0.83, 0.95 and 0.43% for those forest types, respectively. Clearing has declined substantially in the period 2005–09 since the advent of recent regulations in Queensland. The area remains a source of carbon emissions but this situation may reverse if restoration of mulga dry forest becomes an attractive land use with an emerging carbon market.

Eucalyptus melanophloia and E. populnea dominate large areas of savanna in eastern Australia. Under aboriginal management, fires probably occurred under a broad range of conditions, but under pastoral management, burning is avoided when soil moisture is low. This experiment subjected E. melanophloia and E. populnea seedlings to burning and moisture stress, to examine whether this change in burning regime could affect seedling survivorship. The findings suggest survivorship rates are 87–93% for unstressed seedlings with relatively large lignotubers (>12 mm2 plan area) and 56–66% for unstressed seedlings with small lignotubers. There was no substantial interactive effect between moisture stress and burning for E. melanophloia, but such an interaction was apparent for E. populnea, such that moisture stress multiplied the effect of burning. The timing of burning in relation to soil-moisture conditions may have an enduring effect on woodland structure where E. populnea is dominant. E. melanophloia seedlings are more resistant to burning, especially with moisture stress, and fire may not be limiting structural development in woodlands where this species dominates. However, a more detailed understanding of species demography is required, including the conditions required for germination, causes of seedling mortality and the time taken for seedlings to develop fire resistance in the field.

The red-finned blue-eye (Scaturiginichthys vermeilipinnis) is endemic to a single complex of springs emanating from the Great Artesian Basin, Australia. The species has been recorded as naturally occurring in eight separate very shallow (generally

Woody vegetation cover interpreted from aerial photography requires assessment against field data as the signature of woody vegetation cover may differ between photoscales, vegetation types and photo-interpreters. Measurements of aerial woody cover taken from aerial photography of four different photoscales were compared with a field dataset from Eucalyptus- and Acacia-dominated landscapes of semi-arid Queensland. Two interpreters employed a method that utilises a stereoscope and sample-point graticule for manual quantified measurements of aerial woody cover. Both interpreters generated highly significant models accounting for 77 and 78% of deviance. Photoscale appears to have a consistent effect whereby the signature of woody cover increases as the photoscale decreases from 1 : 25 000 to 1 : 80 000, although the magnitude of this effect was different between interpreters. The results suggest no substantial differences in the shape of models predicting crown cover between Acacia- and Eucalyptus-dominated land types, although the precision of the models was greater for the Acacia (90–91% of residual deviance) than for the Eucalyptus (50–56% of residual deviance) land type. The reduced accuracy in the Eucalyptus land type probably reflects the relatively diffuse crowns of the dominant trees. The models generated for this dataset are within the range of those from other calibration studies employing photography of a range of scales and methodologies. The effect of photoscale is verified between the available studies, but there may also be variations arising from methodological differences or image properties. The present study highlights the influence of photoscale and interpreter bias for assessing woody crown cover from aerial photography. Studies that employ aerial photography should carefully consider potential biases and cater for them by calibrating assessments with field measurements.

The effect of ongoing forest invasion on floristic composition of subtropical montane grasslands was considered by examining vascular plant species frequencies across 13 adjoining areas of grassland, invaded grassland (eucalypt forest 50 years old) on the Bunya Mountains in subtropical eastern Australia. Tree invasion of grasslands over the last 50 years has had substantial facilitative or antagonistic net impacts on populations of many plant species. Increases in species frequency, indicating net facilitation, generally appear to occur earlier in forest development than do decreases in frequency, indicative of net antagonism. Although more than 20% of the flora showed substantial association with either grassland or forest, the dominant ground-stratum species in each habitat were quite similar and very few grassland species were not recorded in grassy forests. Forb species composition appears to change more rapidly after tree invasion than grass species composition. Relatively few forbs preferred forest to grassland, whereas shrubs, trees or lianes were substantially more frequent in forests. Replacement of grasslands by grassy forests would reduce landscape diversity and impact on other values in the Bunya Mountains. However, this study suggests that most, if not all, vascular plants that currently occur on the grasslands will persist in the area provided the grassy character of the eucalypt forests is maintained.

Gidgee (Acacia cambagei) forms woodlands and forests on moderately fertile clay soils throughout semi-arid areas of north-eastern Australia. Analysis of soil data from Queensland indicates that gidgee’s environmental domain has a broad overlap with treeless Mitchell grassland although a preference for slightly coarser soils with higher phosphorus and lower potassium content. Mapping of woody cover change for an area of central Queensland suggests that gidgee thickening and spread has occurred between 1951 and 1994. However, this increase has been more than offset by loss of gidgee cover by clearing in other areas. Gidgee thickening and encroachment onto grassland is a problem for graziers because the trees suppress grass and are unpalatable except when very young. Landholder surveys and other information indicate that gidgee has recruited sporadically and that thickening occurred with a massive seedling recruitment in the 1950s with a smaller recruitment wave in the 1970s. These waves coincide with relatively wet times and the extreme rainfall abundance in the 1950s has a multi-century expected recurrence. There are historical records describing gidgee death during the extreme droughts in the first part of the 20th century. Young gidgee plants are fire sensitive but well-established gidgee stands are fire retardant, and the importance of fire for control of gidgee thickening is unknown. One scenario explaining gidgee dynamics emphasises the role of environment (mostly climate and soil), while another emphasises management (mostly fire). Research priorities are presented and may contribute to appropriate management of gidgee woodlands.

The endangered perennial herb Trioncinia retroflexa was re-established within suitable grassland habitat in central Queensland. The trial included direct seeding and transplanting nursery-grown seedlings during the wet season. Successful establishment may be contingent on moist surface soils at the time of planting and the maintenance of this moisture by adequate follow-up rain in the month after planting. Five years after planting there were a total of 22 mature plants and eight infertile plants within two small areas of the re-establishment trial. The established plants are now reproducing and several generations of seedlings have become established. Topographic position (ridges or swales) and post-planting fires had no significant effect on the density of the re-established population. There have been substantial fluctuations in the re-established population in concert with seasonal rainfall patterns. A patchy and small fe-established population could be expected given the patchiness of the largest natural population and the relatively low rainfall of the introduction site.

The vegetation and environmental setting of permanent spring wetlands are described from a survey of 269 spring complexes throughout seasonally arid Queensland. Wetlands associated with springs in the western and southern discharge areas of the Great Artesian Basin are floristically distinct from other spring wetlands. Ordination analysis suggests that the biogeographic regions and the broad geological substrates that support spring wetlands provide a meaningful representation of floristic range. An existing classificatory system that defines ‘regional ecosystems’ on the basis of the biogeographic region and broad geological substrate is adopted to define 15 spring-wetland types in seasonally arid Queensland. The conservation value of the springs is assessed by a scheme that weights plant species populations on the basis of their endemicity and isolation from other populations, demonstrating that both Great Artesian Basin and non-Great Artesian Basin springs have similar conservation values. Australian Journal of Botany 52(5) 583–595

A transect-based quadrat survey was conducted within 11 spring wetlands fed by permanent groundwater flows from the Great Artesian Basin at Elizabeth Springs in western Queensland. Flow patterns within individual wetlands change with sedimentation associated with mound building, siltation of abandoned drains and changes in aquifer pressure associated with artificial extraction from bores. The pattern of floristic groups for the wetland quadrats was poorly related to soil texture, water pH, slope and topographic position. Patterns were most clearly related to wetland age as determined from aerial photography, with a clear successional sequence from mono-specific stands of Cyperus laevigatus on newly formed wetland areas to more diverse wetland assemblages. However, evidence from other Great Artesian Basin springs suggests that succession can also result in reduced species richness where the palatable tall reed Phragmites australis develops mono-specific stands.

Models to calibrate tree and shrub cover assessed from aerial photography with field measurements were developed for a range of vegetation types in north-western Australia. The models verify previous studies indicating that woody cover can be successfully determined from aerial photography. The calibration models were applied to estimates of woody vegetation cover determined for 279 randomly located sample areas in the Ord–Victoria Rivers region using aerial photography from 1948 to 1950 and 1988 to 1997. Overstorey cover increased from a regional average of 11.5% to 13.5% and understorey cover increased from 1.3% to 2.0%. Downs, Limestone Hills and Alluvia land-types showed the most substantial increases in overstorey cover while overstorey cover in the Limestone plains land-type decreased. Relatively open structured vegetation is most susceptible to thickening. Rainfall records reveal an extreme multi-year rainfall deficit in the study area in the 1930s and relatively wet times in the 1970s and 1980s. Interpretation of a limited set of aerial photographs taken between 1964 and 1972 suggests that most of the increases in cover have occurred since this time. The study highlights the possibility that the average trend of vegetation thickening represents recovery during the relatively wet times after the 1970s. There was no relationship between structural change and a grazing intensity surrogate (distance of sample points to stock watering-points). However, the causes of structural change are undoubtedly multi-factored and the relative contributions of climate, fire and grazing vary for different landscapes and tree species.

A nursery experiment revealed that the growth of Eucalyptus tetrodonta seedlings is significantly inhibited by unsterilised monsoon rainforest surface soil, and that this inhibition can not be eliminated by the addition of fertiliser. This result is in marked contrast to similar previously published experiments performed on two rainforest tree seedlings. (Bombax ceiba and Sterculia quadrifida) and Eucalyptus nesophila. The results of this study suggest that the observed high densities of E. tetrodonta seedlings and saplings on a site formerly occupied by monsoon rainforest on Elcho Island could only have established once the rainforest soil biota had been killed. One possible cause of soil sterilisation is wildfire.

Models to calibrate tree and shrub cover assessed from aerial photography with field measurements were developed for a range of vegetation types in north-western Australia. The models verify previous studies indicating that woody cover can be successfully determined from aerial photography. The calibration models were applied to estimates of woody vegetation cover determined for 279 randomly located sample areas in the Ord–Victoria Rivers region using aerial photography from 1948 to 1950 and 1988 to 1997. Overstorey cover increased from a regional average of 11.5% to 13.5% and understorey cover increased from 1.3% to 2.0%. Downs, Limestone Hills and Alluvia land-types showed the most substantial increases in overstorey cover while overstorey cover in the Limestone plains land-type decreased. Relatively open structured vegetation is most susceptible to thickening. Rainfall records reveal an extreme multi-year rainfall deficit in the study area in the 1930s and relatively wet times in the 1970s and 1980s. Interpretation of a limited set of aerial photographs taken between 1964 and 1972 suggests that most of the increases in cover have occurred since this time. The study highlights the possibility that the average trend of vegetation thickening represents recovery during the relatively wet times after the 1970s. There was no relationship between structural change and a grazing intensity surrogate (distance of sample points to stock watering-points). However, the causes of structural change are undoubtedly multi-factored and the relative contributions of climate, fire and grazing vary for different landscapes and tree species.

Studies attempting to calibrate vegetation attributes from aerial photography with field data are reviewed in detail. It is concluded that aerial photography has considerable advantages over satellite-based data because of its capacity to assess the vertical dimension of vegetation and the longer time period the record spans. Limitations of using the aerial photo record as digital data include standardising image contrast and rectification. Some of these problems can be circumvented by manual techniques, but problems of crown exaggeration that varies with photo scale and variation in contrast between the textures of tree crowns and the ground remain. Applications of aerial photography for assessing vegetation change are also reviewed and include deforestation, reforestation, changes in vegetation boundaries, tree density, community composition and crown dieback. These changes have been assessed at scales ranging from individual tree crowns to regional landscapes. In Australia, aerial photography has provided a clear demonstration of deforestation rates and the expansion and contraction of forest and woodland, which is generally attributed to changes in grazing and fire regimes. It is suggested that manual techniques with point-based sampling, digital processing of data for complete spatial coverages and the application of photogrammetric measurements with stereo-plotters are all techniques with great promise for utilising this underrated medium for assessment of vegetation dynamics.

A large floristic data set was collated from vegetation surveys of Mitchell (Astrebla spp.) grasslands in Queensland spanning more than 20 years. The data was ordinated using non-metric multi-dimensional scaling and a four-dimensional solution could be deciphered. The longest floristic gradient was clearly aligned with climatic variables associated with mean annual precipitation. The vector for mean annual temperature was orientated orthogonally to the primary vector and was strongly correlated through the ordination space. A grazing vector tentatively derived from the frequency of species known to be favoured by grazing in Mitchell grasslands was orthogonal to the major climatic trends. However, the suspected grazing influence is complicated by the potential influence of sites occurring on limited areas of limestone habitat representing marginal Mitchell grassland habitat. A vector derived from the binomial categorisation of landform into downs and alluvia was clearly aligned with the fourth dimension of the ordination. Short- and long-term rainfall history variables were generated for each site from local rainfall records. The oblique trajectory of these variables through the ordination space tends to suggest that they may be artefacts of sampling rather than a major independent influence on the character of Mitchell grasslands. Thus the results at the broad scale of this study do not conform to previous studies at the paddock scale that emphasise the influence of climatic fluctuations on the floristic character of Mitchell grasslands. The results suggest Mitchell grasslands will require conservation planning and action throughout their geographic range.

Fifteen references to vegetation structure from the historical land survey record for the eastern Darling Downs were calibrated with historical photographs to determine the use of the surveyors' structural terminology (Open: 'open', 'light': Dense: 'heavy', 'thick', 'dense'). Sites with less than 40% canopy cover were only described by terms included within the Open category, and sites with greater than 50% canopy cover were only described by terms included within the Dense category. These results provide calibrations of the surveyors' structural terminology indicating terms were unambiguously applied outside the 40-50% canopy cover range. The use of corner tree distances as an absolute measure of vegetation density is warned against because corner trees had to meet certain criteria and were not necessarily the nearest tree to any corner. However, the distance from allotment corners to "corner trees" provided a measure of the relative use of the surveyor's structural epithets and is consistent with standard application among their fraternity. Survey records dating from 1864-1910 were compared with the structure of existing remnants (projective crown cover measured from recent 1:25,000 aerial photography) to assess changes in vegetation structure. The analysis suggests that 88% of the 34 sites included in the analysis have not changed from the broad structural category that was assigned by the surveyors. Using the assumptions developed by this study, two sites were assessed as having thickened substantially. These results suggest that only minor vegetation thickening has occurred in the Darling Downs since the early land surveys. This conclusion is supported by direct comparison of the historical photographs with existing remnants. indicating that only one site out of 17 has thickened substantially.

Boggomosses are perennial mound springs in the Dawson River Valley, Queensland, an area subject to seasonal and often more prolonged drought. The results of a vegetation survey of the boggomosses are presented and assessed in relation to an impoundment proposed for the area. A classification of site-species data defined four boggomoss communities having distinct associations with soil texture and fertility, landscape position and possibly the age of springs. A measure of biogeographic significance was assigned to the plant species on the basis of rarity, isolation, affinity with the coastal flora and latitudinal limits. An impoundment at the maximum proposed height of 185 m would: a) inundate about 58% of the total number of springs; b) inundate 62% of significant community 1 springs that have high conservation significance; c) inundate all boggomoss populations of 26% of native plant species; d) inundate 30% of the boggomoss populations of the species that currently have two or more boggomoss populations to a level of one or less boggomoss populations; e) inundate all boggomoss populations of two out of 25 biogeographically significant species; f) inundate six out of 12 boggomoss populations of the vulnerable species Arthraxon hispidus; g) not inundate the single population of the vulnerable species Thelypteris confluens although the population would be within 1 m altitude of the maximum water height; h) not inundate the two populations of the endangered species Eriocaulon carsonii and Myriophyllum sp. (Aramac B. Wilson 110). Substantial lessening of impact on community 1 sites are achieved at dam water levels down to 177 m altitude and this trend is reftected in a progessive increase in the security of individual species. For example at the latter level, 88% of species would remain intact in more than one population and all known populations of Arthraxonwould remain intact.

Preferential clearance of productive Acacia-dominated ecosystems was determined using digital land-system mapping and a 1994–1995 native vegetation remnant coverage from central Queensland. The previously mapped land systems are composed of complexes of land units, covering a range of productive potentials that are not individually mapped but defined according to their proportions within the land systems. Some land units were grouped into 11 ‘Acacia ecosystems’ composed of unique combinations of geology and vegetation dominated by brigalow (Acacia harpophylla F.Muell. ex Benth.), gidgee (A. cambagei R.T.Baker) or blackwood (A. argyrodendron Domin). Calculation of the remnant area of these Acacia ecosystems, assuming equal proportions of vegetation clearance of the land unit components within the land-system complexes (equal proportion calculation), was compared with a more accurate calculation based on interpretation of aerial photography (aerial-photograph calculation) data. The aerial photograph calculation of remnant area was 44% of the equal proportion calculation for brigalow on Tertiary clay and the ratio between these calculations was greater than 100% for only one Acacia ecosystem of relatively low production potential. The proportion of the remnant Acacia ecosystems relative to less productive ecosystems on tenures assigned to production (leasehold and freehold) was consistently lower than on other tenures. The use of mapping consisting of composite ecosystems with various production potentials for area-based assessment of vegetation clearance will overestimate the remnant areas of the relatively productive types unless a procedure such as that outlined here is used to correct for preferential clearance. This study provides near-complete and accurate data on the 1994–1995 status of Acacia ecosystems in central Queensland. Relative to their original area, values ranged from 6.8% for brigalow on Tertiary clay to 37.6% for gidgee–blackwood on alluvium.

The moss flora of dry rainforest in north Queensland was surveyed in conjunction with a vascular plant survey. Moss species richness was strongly correlated with patch area mean annual rainfall, and vascular plant species richness. Moss species richness rises where volcanic craters within the study area increase moisture status. Geological substrate had a stronger relationship with associations within the mass flora as defined by the TWINSPAN classificatory procedure and with individual moss species than either landform category or classificatory groups of vascular plants. The association of four species with large closed canopy patches may reflect a humid microclimate and suggests that all else being equal large patches will preserve more species than many small areas of rainforest. However, a high proportion of rare species were associated with the soil surface and there was no obvious way of predicting their locality.

The physical setting of 61 grassy balds on the Bunya Mountains in south-eastern Queensland was surveyed during 1995, and a further 73 balds were assessed from aerial photographs taken in 1951 and 1991. Balds were analysed to search for consistent environmental patterns that may account for their existence. There was little association between measured environmental variables (including altitude, surface rock cover, slope, landscape position, surrounding vegetation, bald area) and the occurrence of balds, although large balds have a weak association with steep slopes and relatively low rock cover. Grassland was on shallower soil than forest. Some, but not all balds were surrounded by rainforest. No consistent relationship was evident for all measured balds surrounded by eucalypt forest. About 26% of the area of balds existing in 1951 had been invaded by forest by 1991. The extent of the invasion was generally higher for balds surrounded by eucalypt forest than balds surrounded by eucalypt and rainforest or rainforest only, but showed little relationship with other environmental variables. The invading tree species groups could not be associated with environmental variables. The lack of environmental correlation of bald location and size, and the rapid change in bald area is not supportive of hypotheses relating bald maintenance to intrinsic physical conditions. There is relatively poor species diversity of Eucalyptus on the Bunya Mountains and this narrow ecological range of species may be a limitation on the colonisation of the grasslands. Indirect evidence for the role of grazing and fire as agents of bald maintenance are discussed, and it is concluded, that fire, and not grazing, is likely to be a major influence.

Site-species data were collected from inland dry rainforest of north Queensland. The extent of patches declines with distance from the coast and they do not occur where rainfall is less than about 500mm per annum. TWINSPAN derived eight broad groups and sixteen sub- groups of the floristic continuum of rainforest in the study area. The floristic sub-groups are restricted within regions of the study area and most occur over a wide range of geological substrates and landforms. The primary floristic gradient is related to a latitudinal trend coinciding with the coefficient of variation of monthly precipitation, and a secondary gradient to distance from the coast and mean annual precipitation. The placement of dry rainforest from the study area within existing schemes is examined and the past terminology of Australian rainforest from relatively dry environments is discussed. Nearly half of the dry rainforest flora is obligately associ- ated with that environment. Although the flora is fairly spe- cialized to the rainforest environment, 87% of the inland woody species inventoried during this study are also known to occur within 1 km of the coast. It is argued that the wide- spread nature of the flora, the lack of specialization to habitat, and the existence of relatively complex dry rainforest on basalt lava flows of late Pleistocene age are consistent with a relatively recent invasion of this region of Queensland by this vegetation type since drier times during the last Glacial cycle.

Radiocarbon dates confirm a chronological sequence for late Holocene beach ridges at Wangiti Beach on Bathurst Island. The vegetation on these beach ridges can be clearly related to topography and distance from the sea. Monsoon rainforest occurs on the fore-dunes where the nutrient levels of the young sediments are relatively high. The mid-dunes support woodland dominated by Melalueca viridiflora and have nutrient-poor acid soils. Vegetation with a high component of monsoon rainforest species occupies the rear dunes, which have a high concentration of seawater macronutrient cations despite their older age than more seaward dunes. The relatively low elevation of the rear dunes supports the suggestion that soil nutrient levels are enriched by sea water or marine sediments during infrequent inundation events such as those during cyclones.

Downslope boundaries of forest with grassland and grassy woodland occur over a wide altitudinal range in central Tasmania. Three sites were selected for study of the causes of these boundaries at low, medium and high elevations. The open vegetation was generally associated with moister and less rocky soils and more subdued topography than the adjacent forest. Frost incidence and intensity, soil moisture and waterlogging varied markedly among the three open areas. Planted tree seedlings survived 4 years in the open at all sites, and seedlings established in the open both naturally, and after sowing, where grass competition was reduced by herbicide application, digging or root competition from adult eucalypts. Grazing had no detectable effect on seedling establishment. A pot experiment demonstrated a suppressive effect of native grass swards on both seedling establishment and growth, this effect being largely independent of available moisture and nutrients. While frost, waterlogging, fire and drought may play a role in inhibiting eucalypt establishment and increasing eucalypt mortality at some or all of the sites, the dense grass swards found in all the open areas are considered to be the most likely primary agent of tree exclusion.

The stand structure of tree species in tropical eucalypt forest on Melville Island reveals a mass of short woody sprouts in the ground layer and low numbers of sapling eucalypts. The growth of the woody sprouts showed no significant response in the first 2 years after release from overwood competition. However, eucalypts are released in response to overwood removal, after 2-5 years, although investigations of old clear-felled blocks indicated that this response is not consistent. The initiation of saplings may be related to the size of the lignotuber and the presence of a tap root for some species. It is suggested that the accession of saplings may be limited by the degradation of root systems by termite herbivory. Using assumptions regarding longevity of life stages, it is demonstrated that the forest structure of the study site can be perpetuated undercurrent conditions despite indications that the relative dominance of the forest eucalypt species will shift.

In the savannah of Melville Island, Northern Temtory floristic variation relates strongly to both soil moisture regimes at different depths in the profile and chemical edaphic conditions. While some elements of variation in vegetation structure and overstorey species composition have clear edaphic relationships, many common tree species could be found as important overstorey components in a range of edaphic situations. The interaction of stochastic disturbance events and phenological condition may explain this deviation from environmental determinism.

A model of direct and diffuse solar radiation on slopes of varying angle and aspect suggests that radiation differences within hilly terrain are maximized in the beginning of the dry season in the monsoon tropics. The differences between north and south facing slopes are greater than those found during the summer in the temperate zone. Within a study area near Mt. Bundey, Northern Territory, floristic and structural vegetational variability is closely related to June radiation as estimated by the model. However, the distribution patterns of monsoon thicket and eucalypt forest relate more to relative five protection than to the effects of incident radiation on temporal patterns of moisture availability. Within both major formations, site rockiness is an important influence on vegetation floristics and structure, both for five protection and through its influence on moisture availability.

Most studies of climate change effects on fire regimes assume a gradual reorganization of pyrogeographic patterns and have not considered the potential for transformational changes in the climate-vegetation-fire relationships underlying continental-scale fire regimes. Here, we model current fire activity levels in Australia as a function of mean annual actual evapotranspiration (E) and potential evapotranspiration (E 0), as proxies for fuel productivity and fuel drying potential. We distinguish two domains in $E,{E}_{0}$ space according to the dominant constraint on fire activity being either fuel productivity (PL-type fire) or fuel dryness (DL-type fire) and show that the affinity to these domains is related to fuel type. We propose to assess the potential for transformational shifts in fire type from the difference in the affinity to either domain under a baseline climate and projected future climate. Under the projected climate changes potential for a transformational shift from DL- to PL-type fire was predicted for mesic savanna woodland in the north and for eucalypt forests in coastal areas of the south–west and along the Continental Divide in the south–east of the continent. Potential for a shift from PL- to DL-type fire was predicted for a narrow zone of eucalypt savanna woodland in the north–east. This research was partly financially supported by the Australian Centre for Ecological Analysis and Synthesis (ACEAS). VRD was partly funded by a Ramón y Cajal Fellowhip (RYC-2012-10970). Field data collection and MAC were financially supported by NASA Interdisciplinary Sciences Grant (NNX11AB89G).

Some 83% of the area of native vegetation of the Tasmanian Midlands has been replaced after 170 yr of agricultural exploitation, but scattered remnants of the native vegetation still exist. The floristic variation within and between these remnants was most closely related to soil moisture, as affected by texture and topography. A classification produced 13 groups that comprised subdivisions of grassy communities dominated by Eucalyptus amygdalina forest, E. viminalis, E. pauciflora, E. ovata woodlands and treeless communities. From accounts of surveyors and travellers, survey plans, paintings and standing paddock trees, it was determined that these formations dominated the Midlands at the time of European contact. -from Author