Extent and timing of floristic exchange between Australian and Asian rain forests

Aim We tested an entrenched concept – that the Australian rain forest flora is essentially a Gondwanan relict. We also assessed the role of regional‐level source–sink dynamics in the assembly of this flora. Location Eastern Australia. Methods To avoid potential biases inherent in selective studies undertaken to date, we used an analytical, whole‐of‐flora approach integrated with the fossil record. We identified disjunctions between woody Australian rain forest plant taxa and relatives on other land masses. To test the strength of the fossil evidence for the regional antiquity of this flora, we evaluated the proportion of these disjunct clades represented in the Australian fossil record, and to minimize the effects of biases in this record, we compared late Quaternary (i.e. late Pleistocene and Holocene, 126–0 ka), Pliocene and late Oligocene–early Miocene Australian pollen records interpreted as tropical rain forest. Using within‐species disjunctions as a proxy, we assessed the role of recent immigration from Asia into Australia. To assess the role of source–sink dynamics, we performed comparative analyses of disjunctions in major rain forest categories representing a north–south/climatic gradient. Results Southern Australian, cool temperate (microthermal) rain forests contain many floristic disjunctions with Gondwanan fragments and most of these clades have Gondwanan fossils. Disjunct clades in Australian mesothermal rain forest mostly occur in Asia/Malesia and a low proportion of these clades show pre‐Neogene records. Many clades in lowland tropical and ‘dry’ rain forest show disjunctions with Asia/Malesia and few have Australian fossil records. Rates of recent immigration from Asia/Malesia are high in these northern forests, and outweigh rates of recent emigration approximately nine‐fold. The late Quaternary fossil record has many more rain forest angiosperms than Oligocene–Miocene and Pliocene floras, consistent with extensive late Cenozoic immigration. Main conclusions The microthermal rain forests are largely Gondwanan relicts, but there is progressively greater, and more recent contribution from Asia/Malesia into more northern, and more lowland tropical rain forests. This variation reflects a strong gradient in geographic and ecological proximity between these forests and source floras in Asia/Malesia, and is consistent with a source–sink size model of immigration driven by late Cenozoic contractions and expansions of Australian rain forest.     

Evidence of a Holocene and continuing recent expansion of lowland rain forest in humid, tropical North Queensland

Abstract. The identification and radiocarbon dating of charcoal collected under tropical rain forest indicated that sclerophyll forests dominated by Eucalyptus occupied parts of the wet tropical lowlands in the Daintree region of North Queensland at least intermittently from 12,000 yr bp until very recently. The results extend the late Pleistocene expansion of pyrophytic, sclerophyll forests which occurred in the upland rain forests to a humid, megathermic coastal lowlands region. Unlike the early Holocene re-expansion of rain forests which occurred generally on the uplands, the sclerophyll forests in the lowland study area were present until at least 1400 yr bp. Changes in coastal geomorphology and coastline positions during the late Quaternary were examined in the study area by superimposing sea levels derived from published curves on sea-bed contours. The results indicate that a very rapid decrease in the extent of the coastal plain occurred during the late Pleistocene. Between 12,000 and 9000 yr bp, 26 km of the coastal plain was submerged and this would have inevitably resulted in concentrations of Aboriginal populations in the area of the present coastline. It is suggested that burning activities by Aborigines in the coastal lowlands were sufficient to reestablish sclerophyll forests during the latter part of the Holocene from approximately 4000 yr bp following a wanner and wetter period which would have been conducive to rain forest re-expansion. Although the evidence suggests that the most recent rain forest recolonization occurred in the study area more than 1000 years ago, the process is still continuing elsewhere in the wet tropical lowlands in North Queensland. The process of eucalypt forest replacement by rain forest may have accelerated since the arrival of Europeans and the concomitant decrease in Aboriginal management.     

Pollen-based reconstructions of biome distributions for Australia, Southeast Asia and the Pacific (SEAPAC region) at 0, 6000 and 18,000 14C yr BP

Aim This paper documents reconstructions of the vegetation patterns in Australia, Southeast Asia and the Pacific (SEAPAC region) in the mid‐Holocene and at the last glacial maximum (LGM). Methods Vegetation patterns were reconstructed from pollen data using an objective biomization scheme based on plant functional types. The biomization scheme was first tested using 535 modern pollen samples from 377 sites, and then applied unchanged to fossil pollen samples dating to 6000 ± 500 or 18,000 ± 1000 ¹⁴C yr bp . Results 1. Tests using surface pollen sample sites showed that the biomization scheme is capable of reproducing the modern broad‐scale patterns of vegetation distribution. The north–south gradient in temperature, reflected in transitions from cool evergreen needleleaf forest in the extreme south through temperate rain forest or wet sclerophyll forest (WSFW) and into tropical forests, is well reconstructed. The transitions from xerophytic through sclerophyll woodlands and open forests to closed‐canopy forests, which reflect the gradient in plant available moisture from the continental interior towards the coast, are reconstructed with less geographical precision but nevertheless the broad‐scale pattern emerges. 2. Differences between the modern and mid‐Holocene vegetation patterns in mainland Australia are comparatively small and reflect changes in moisture availability rather than temperature. In south‐eastern Australia some sites show a shift towards more moisture‐stressed vegetation in the mid‐Holocene with xerophytic woods/scrub and temperate sclerophyll woodland and shrubland at sites characterized today by WSFW or warm‐temperate rain forest (WTRF). However, sites in the Snowy Mountains, on the Southern Tablelands and east of the Great Dividing Range have more moisture‐demanding vegetation in the mid‐Holocene than today. South‐western Australia was slightly drier than today. The single site in north‐western Australia also shows conditions drier than today in the mid‐Holocene. Changes in the tropics are also comparatively small, but the presence of WTRF and tropical deciduous broadleaf forest and woodland in the mid‐Holocene, in sites occupied today by cool‐temperate rain forest, indicate warmer conditions. 3. Expansion of xerophytic vegetation in the south and tropical deciduous broadleaf forest and woodland in the north indicate drier conditions across mainland Australia at the LGM. None of these changes are informative about the degree of cooling. However the evidence from the tropics, showing lowering of the treeline and forest belts, indicates that conditions were between 1 and 9 °C (depending on elevation) colder. The encroachment of tropical deciduous broadleaf forest and woodland into lowland evergreen broadleaf forest implies greater aridity. Main conclusions This study provides the first continental‐scale reconstruction of mid‐Holocene and LGM vegetation patterns from Australia, Southeast Asia and the Pacific (SEAPAC region) using an objective biomization scheme. These data will provide a benchmark for evaluation of palaeoclimate simulations within the framework of the Palaeoclimate Modelling Intercomparison Project.     

New approaches to understanding late Quaternary climate fluctuations and refugial dynamics in Australian wet tropical rain forests

Aim We created spatially explicit models of palaeovegetation stability for the rain forests of the Australia Wet Tropics. We accounted for the climatic fluctuations of the late Quaternary, improving upon previous palaeovegetation modelling for the region in terms of data, approach and coverage of predictions. Location Australian Wet Tropics. Methods We generated climate‐based distribution models for broad rain forest vegetation types using contemporary and reconstructed ‘pre‐clearing’ vegetation data. Models were projected onto previously published palaeoclimate scenarios dating to c. 18 kyr bp . Vegetation stability was estimated as the average likelihood that a location was suitable for rain forest through all climate scenarios. Uncertainty associated with model projections onto novel environmental conditions was also tracked. Results Upland rain forest was found to be the most stable of the wet forest vegetation types examined. We provide evidence that the lowland rain forests were largely extirpated from the region during the last glacial maximum, with only small, marginally suitable fragments persisting in two areas. Models generated using contemporary vegetation data underestimated the area of environmental space suitable for rain forest in historical time periods. Model uncertainty resulting from projection onto novel environmental conditions was low, but generally increased with the number of years before present being modelled. Main conclusions Climate fluctuations of the late Quaternary probably resulted in dramatic change in the extent of rain forest in the region. Pockets of high‐stability upland rain forest were identified, but extreme bottlenecks of area were predicted for lowland rain forest. These factors are expected to have had a dramatic impact on the historical dynamics of population connectivity and patterns of extinction and recolonization of dependent fauna. Finally, we found that models trained on contemporary vegetation data can be problematic for reconstructing vegetation patterns under novel environmental conditions. Climatic tolerances and the historical extent of vegetation may be underestimated when artificial vegetation boundaries imposed by land clearing are not taken into account.     

Decline of a biome: evolution,contraction, fragmentation,extinction and invasion of the Australian mesic zone biota

Aim The mesic biome, encompassing both rain forest and open sclerophyllous forests, is central to understanding the evolution of Australia’s terrestrial biota and has long been considered the ancestral biome of the continent. Our aims are to review and refine key hypotheses derived from palaeoclimatic data and the fossil record that are critical to understanding the evolution of the Australian mesic biota. We examine predictions arising from these hypotheses using available molecular phylogenetic and phylogeographical data. In doing so, we increase understanding of the mesic biota and highlight data deficiencies and fruitful areas for future research. Location The mesic biome of Australia, along the eastern coast of Australia, and in the south‐east and south‐west, including its rain forest and sclerophyllous, often eucalypt‐dominated, habitats. Methods We derived five hypotheses based on palaeoclimatic and fossil data regarding the evolution of the Australian mesic biota, particularly as it relates to the mesic biome. We evaluated predictions formulated from these hypotheses using suitable molecular phylogenies of terrestrial plants and animals and freshwater invertebrates. Results There was support for the ancestral position of mesic habitat in most clades, with support for rain forest habitat ancestry in some groups, while evidence of ancestry in mesic sclerophyllous habitats was also demonstrated for some plants and herpetofauna. Contraction of mesic habitats has led to extinction of numerous lineages in many clades and this is particularly evident in the rain forest component. Species richness was generally higher in sclerophyllous clades than in rain forest clades, probably due to higher rates of net speciation in the former and extinction in the latter. Although extinction has been prominent in rain forest communities, tropical rain forests appear to have experienced extensive immigration from northern neighbours. Pleistocene climatic oscillations have left genetic signatures at multiple levels of divergence and with complex geographical structuring, even in areas with low topographical relief and few obvious geographical barriers. Main conclusions Our review confirms long‐held views of the ancestral position of the Australian mesic biome but also reveals new insights into the complexity of the processes of contraction, fragmentation, extinction and invasion during the evolution of this biome.     

Firescape ecology: how topography determines the contrasting distribution of fire and rain forest in the south‐west of the Tasmanian Wilderness World Heritage Area

Aim To test the hypothesis that ‘islands’ of fire‐sensitive rain forest are restricted to topographic fire refugia and investigate the role of topography–fire interactions in fire‐mediated alternative stable state models. Location A vegetation mosaic of moorland, sclerophyll scrub, wet sclerophyll eucalypt forest and rain forest in the rugged, fire‐prone landscapes of south‐west Tasmania, Australia. Methods We used geospatial statistics to: (1) identify the topographic determinants of rain forest distribution on nutrient‐poor substrates, and (2) identify the vegetation and topographic variables that are important in controlling the spatial pattern of a series of very large fires (> 40,000 ha) that were mapped using Landsat Thematic Mapper (TM) satellite imagery. Results Rain forest was more likely to be found in valleys and on steep south‐facing slopes. Fires typically burned within highly flammable treeless moorland and stopped on boundaries with less flammable surrounding vegetation types such as wet sclerophyll forest and rain forest. Controlling for the effect of vegetation, fires were most likely to burn on flats, ridges and steep north‐facing slopes and least likely to burn in valleys and on steep south‐facing slopes. These results suggest an antagonism between fire and rain forest, in which rain forest preferentially occupies parts of the landscape where fire is least likely to burn. Main conclusions The distribution of rain forest on nutrient‐poor substrates was clearly related to parts of the landscape that are protected from fire (i.e. topographic fire refugia). The relative flammability of vegetation types at the landscape scale offers support to the proposed hierarchy of fire frequencies (moorland > scrub > wet sclerophyll > rain forest) that underpins the ecological models proposed for the region. The interaction between fire occurrence and a range of topographic variables suggests that topography plays an important role in mediating the fire–vegetation feedbacks thought to maintain vegetation mosaics in south‐west Tasmania. We suggest that these fire–topography interactions should be included in models of fire‐mediated alternative stable vegetation states in other fire‐prone landscapes.     

Geographic variation in the ecological effects of extinction of Australia's Pleistocene megafauna

Recent studies suggest that extinction of Pleistocene megafauna had large impacts on the structure and functioning of ecosystems, including increased fire and shifts in vegetation state. We argue that the ecological effects of mega‐herbivore extinction are likely to have varied geographically, and might have been reduced in environments of low productivity. We tested this at Caledonia Fen, a cool, high‐elevation site in southeast Australia with a palynological record reaching back approximately 140 ka. The dung fungus Sporormiella indicated that large herbivores were present through most of the early part of the last glacial cycle, but declined abruptly between 50–40 ka and did not recover. This event corresponds with evidence for continent‐wide extinction of Australia's Pleistocene megafauna at that time. An earlier episode of low Sporormiella occurrence coincided with evidence of raised water levels in the fen. Changes in wetland conditions can alter the accumulation of Sporormiella, but there was no such change when Sporormiella counts fell in the period 50–40 ka. We found no evidence that the decline in Sporormiella triggered increased fire or a change in vegetation, which remained a low grass/shrub steppe. This contrasts with a warmer and more humid site, Lynch's Crater in northeast Australia, where decline of dung fungi was followed by increased fire and transition from mixed sclerophyll forest and rainforest to uniform sclerophyll forest. Our results suggest that the magnitude of ecological responses to Pleistocene megafaunal extinction varied geographically, under the control of regional climates.     

Rain forest invasion of eucalypt-dominated woodland savanna, Iron Range, north-eastern Australia: II. Rates of landscape change

Aim To explore rates of rain forest expansion and associated ecological correlates in Eucalyptus‐dominated woodland savanna vegetation in north‐eastern Australia, over the period 1943–91. Location Iron Range National Park and environs, north‐east Queensland, Australia. This remote region supports probably the largest extent of lowland (< 300 m) rain forest extant in Australia. Rainfall (c. 1700 mm p.a.) occurs mostly between November and June, with some rain typically occurring even in the driest months July–October. Methods Interpretation of change in lowland rain forest vegetation cover was undertaken for a 140 km² area comprising complex vegetation, geology and physiography using available air photos (1943, 1970 and 1991). A GIS database was assembled comprising rain forest extent for the three time periods, geology, elevation, slope, aspect, proximity to streams and roads. Using standard GIS procedures, a sample of 6996 10 × 10 m cells (0.5% of study area) was selected randomly and attributed for vegetation structure (rain forest and non‐rain forest), and landscape features. Associations of rain forest expansion with landscape features were examined with logistic regression using the subset of cells that had changed from other vegetation types to rain forest, and remained rain forest over the assessment period, and comparing them with cells that showed no change from their original, non‐rain forest condition. Results Rain forest in the air photo study area increased from 45 km² in 1943 to 78.1 km² by 1970, and to 82.6 km² by 1991. Rainfall (and atmospheric CO₂ concentration) was markedly lower in the first assessment period (1943–70). Modelled rates of rain forest invasion differed predominantly with respect to substrate type, occurring faster on substrates possessing better moisture retention properties, and across all elevation classes. Greatest expansion, at least in the first assessment period, occurred on the most inherently infertile substrates. Expansion was little constrained by slope, aspect and proximity to streams and roads. On schist substrates, probability of invasion remained high (> 60%) over distances up to 1500 m from mature rain forest margins; on less favourable substrates (diorite, granites), probability of expansion was negligible at sites more than 400 m from mature margins. Main conclusions (i) Rain forest expansion was associated primarily with release from burning pressure from c. the 1920s, following major disruption of customary Aboriginal lifestyles including hunting and burning practices. (ii) Decadal‐scale expansion of rain forest at Iron Range supports extensive observations from the palaeoecological literature concerning rapid rain forest invasion under conducive environmental conditions. (iii) The generality of these substrate‐mediated observations requires further testing, especially given that landscape‐scale rain forest invasion of sclerophyll‐dominated communities is reported from other regions of north‐eastern Australia.     

Rain forest invasion of eucalypt-dominated woodland savanna, Iron Range, north-eastern Australia: I. Successional processes

Aim To explore successional processes associated with rain forest expansion in Eucalyptus‐dominated woodland savanna vegetation in north‐eastern Australia. Location Iron Range National Park and environs, northeast Queensland, Australia. This remote region supports probably the largest extent of lowland (< 300 m) rainforest remnant in Australia. Rainfall (c. 1700 mm p.a.) occurs mostly between November and June, with some rain typically occurring even in the driest months July–October. Methods (1) Sampling of rain forest seedling distributions, and other vegetation structural attributes, in fifteen 10 × 10 m quadrats distributed equi‐distantly between mature rain forest margins (range: 70–840 m), at each of 10 sites which were open‐canopied vegetation in 1943. (2) Assessment of relationships between rain forest seedling densities and structural characteristics, including distance‐to‐rain forest‐margin, canopy height, stem density. (3) Assessment of lifeform and dispersal spectra for defined vegetation structural types. Results Rates of rain forest invasion were found to be substrate‐mediated. Transects established on hematite schist, diorite, riverine alluvium, and granite developed closed canopies (termed phase III sites) by 1991. The remainder (four transects on poorly drained colluvial/alluvial sediments; one on dune sands) continued to occur either as grassy woodland (phase I), or with developing rain forest understoreys (phase II). Rain forest seedlings were observed at maximum sampled distances from mature rain forest margins at all sites. Lifeform and dispersal spectra data illustrated that: (1) the proportions of woodland trees, shrubs and graminoids declined with successional phase, with concomitant increases in rain forest primary trees and all other lifeform categories save rain forest trees; (2) the proportions of major dispersal syndromes did not vary between successional phases, neither for rain forest nor woodland taxa. Main conclusions Rain forest seedling distribution data for phases I and II sites illustrate three successional processes: margin extension – seedling density significantly negatively correlated with distance from mature rain forest margins at two sites; nucleation – seedling densities significantly positively correlated with tall trees at two sites; and irruption – seedling densities at two sites neither correlated with distance from mature rain forest margins, nor with measured vegetation structural features. The observation of irruptive rain forest regeneration at these sites, combined with decadal‐scale rain forest canopy development at the five remaining sites, illustrates that under conditions conducive to growth (moisture, substrate), low fire disturbance, and maintenance of diverse dispersal processes (high frugivore richness), rain forest can rapidly invade regional landscapes.     

A fine-resolution Pliocene pollen and charcoal record from Yallalie, south-western Australia

Aim This paper aims to reconstruct a high‐resolution fire and vegetation history from a period when humans were absent in Australia. This is then used to comment on the frequency of natural fire in high biodiversity heathland, and to compare this with historical fire regime in the same region. Methods A section of varved sediment covering a period of c. 84 years was taken from Palaeolake Yallalie in south‐western Australia. The sediments were separated into approximately single to small multiples of years and then analysed for charcoal, pollen and sediment analysis to reconstruct the environmental conditions at the time. Results The charcoal record indicates fire recurrence to have been roughly between 5 and 13 years, a little longer than those of the historical period. The pollen record was dominated by Casuarinaceae, Myrtaceae and a large number of Proteaceae species; these are intermixed with Araucariaceae, Nothofagus and Podocarpus. This suggests there was a mix of sclerophyll woodland and a mosaic of rain forest elements, thus conditions must have been wetter, particularly in the summers, compared with today. Conclusions We assume that fire was most likely confined to the sclerophyll vegetation, and that fire has been a significant feature of the environment long before humans entered Australia. The slightly longer fire recurrence times compared with the present result from the intermittent nature of lightning and wetter summers at the time.     

Classification,species richness,and environmental relations of monsoon rain forest in northern Australia

A floristic classification for monsoon rain forest vegetation in the Northern Territory, Australia, is derived based on comprehensive floristic inventory and environmental data. Allied aims include relating the floristic classification to Australia-wide structural and floristic schema, documenting species richness, and exploring site-environmental relations. TWINSPAN classification and complementary DCA analysis of a data set comprising 1219 sites x 55 9 rain forest taxa yielded 16 floristic assemblages. A diagnostic floristic key to these groups is presented. Eight groups describe rain forests associated with sites of perennial moisture; eight groups are associated with seasonally dry landforms. The structural typology of Australian rain forests is found wanting when applied to relatively simple monsoon rain forest communities. Rain forest patches are mostly less than 5 ha in size; maximum species richness is ca. 135 species per patch. Two major environmental gradients are identified through indirect gradient analysis: a primary latitudinal-moisture gradient and a subsidiary topographic-drainage gradient. Given the demonstrated tolerance of monsoon rain forest to a broad range of environmental conditions, the question remains: why is this vegetation type so restricted in occurrence in northern Australia?     

Temperate forests of the southern hemisphere

The much higher proportion of sea to land in the southern hemisphere creates conditions favouring temperate rain forest on west-facing coasts, so that forests dominated by Nothofagus spp. and southern conifers in Araucariaceae and Podocarpaceae have survived since Cretaceous times. By contrast only Australia preserves to any significant extent forests of a xeromorphic character in drier areas. During the progressive desiccation of the Tertiary era, the prevalent nutrient deficiency in Australian soils led to the evolution of xeromorphic forests of a pronouncedly sclerophyll type in which many elements were inherently fire-resistant. Such forests were enabled to withstand the intensification of burning following the arrival of early man, with selection for the more highly fire-resistant elements such as Eucalyptus which became widely dominant. In other land areas of the southern hemisphere fire has largely destroyed the drier forests.     

Assessing regulating and provisioning ecosystem services in a contrasting tropical forest landscape

Ecosystem services are the bridge between nature and society, and are essential elements of community well-being. The Wet Tropics Australia, is environmentally and biologically diverse, and supplies numerous ecosystem services. It contributes to the community well-being of this region, Australian national economy and global climate change mitigation efforts. However, the ecosystem services in the region have rarely been assessed undermining strategic landscape planning to sustain their future flow. In this study, we attempted to: (i) assess the quantity of five regulating ecosystem services – global climate regulation, air quality regulation, erosion regulation, nutrient regulation, and cyclone protection, and three provisioning ecosystem services – habitat provision, energy provision and timber provision across rainforests, sclerophyll forests and rehabilitated plantation forests; (ii) evaluate the variation of supply of those regulating and provisioning ecosystem services across environmental gradients, such as rainfall, temperature, and elevation; (iii) show the relationships among those ecosystem services; and (iv) identify the hotspots of single and multiple ecosystem services supply across the landscape. The results showed that rainforests possess a very high capacity to supply single and multiple ecosystem services, and the hotspots for most of the regulating and provisioning ecosystem services are found in upland rainforest followed by lowland rainforest, and upland sclerophyll forest. Elevation, rainfall and temperature gradients along with forest structure are the main determinant factors for the quantity of ecosystem services supplied across the three forest types. The correlation among ecosystem services may be positive or negative depending on the ecosystem service category and vegetation type. The rehabilitated plantation forests may provide some ecosystem services comparable to the rainforest. The results demonstrated disturbance regimes (such as tropical cyclones) may have influenced the usual spatial trend of ecosystem service values. This study will assist decision makers in incorporating ecosystem services into their natural resource management planning, and for practitioners to identify the areas with higher values of specific and multiple ecosystem services.     

The value of structural features in tropical forest typology

A pro forma is presented which is based on selected physiognomic–structural features of tropical forest vegetation. It was designed to facilitate collection of data by non-botanists, and the occurrence of features was assessed either by presence/absence or on a four-point scale. Data were collected in a mixture of rain forests and sclerophyll forests in North Queensland and New Guinea, and analysed by an agglomerative polythetic method to provide a classification of sites. The structural classification of the simple forests (regrowth rain forests, forests under extreme limiting conditions and eucalypt woodlands) was unsatisfactory, and floristic differentiation is necessary. However, the structural classification of the complex forests, and of mature forest vegetation over a wide environmental range, produced meaningful site-groupings below the subformation level. The method of structural analysis is rapid and relatively inexpensive, and shows most promise in complex tropical vegetation for which floristic data are not usually available.     

Cyclic habitat displacements during Pleistocene glaciations have induced independent evolution of Tasimia palpata populations (Trichoptera: Tasimiidae) in isolated subtropical rain forest patches

Aim Mechanisms generating biodiversity and endemism are influenced by both historical and ecological patterns, and the relative roles of history vs. ecological interactions are still being debated. The phylogeography of one rain forest‐restricted caddisfly species, Tasimia palpata, thought to have good dispersal abilities, is used to address questions about shifts of highland rain forest habitat during Pleistocene glaciations and about their consequences for haplotype composition and distribution. Location Tasimia palpata occurs in highland subtropical rain forest patches, which are separated from one another by lowland dry bush, in south‐eastern Queensland, Australia. Methods We sequenced 375 base pairs of the mitochondrial cytochrome oxidase I gene from 169 individuals (20 populations) of T. palpata, mainly from three fragmented subtropical rain forest blocks, revealing 46 haplotypes. Analysis of molecular variance (amova ), genetic divergence between populations, nested clade analyses and tests based on coalescent theory were used to analyse phylogeographical relationships among T. palpata populations. Results amova indicates spatial genetic structure between isolated subtropical rain forest patches, with an isolation‐by‐distance effect. Tests based on coalescent theory suggest a repeated process of population reductions and divergence between isolated rain forests during Pleistocene glaciations as a consequence of habitat constrictions followed by population expansions during interglacial periods when subtropical rain forest expanded. In addition, these results suggest that, prior to the Pleistocene, rain forest and T. palpata had more widespread distributions in this region. Main conclusions Historical rain forest expansion and contraction during the Pleistocene resulted in changes in demography and genetic diversity of T. palpata, as well as in an increase in genetic divergence between populations from different patches of subtropical rain forest. Despite the fact that this caddisfly species was isolated in separate highland rain forest patches at various times during the Pleistocene, there is no evidence of allopatric speciation during the Quaternary, which contrasts with other examples of endemism and high diversity in rain forest highlands.     

Fire persistence traits can be used to predict vegetation response to changing fire regimes at expansive landscape scales – an Australian example

Aim Building on a substantial literature addressing the fire responses of woody plants, particularly under mediterranean climates, we assess the extent to which fire persistence traits can be used to predict vegetation responses to fire regime changes in fire‐prone arid and savanna landscape settings. Location Australia, applying data from arid central to monsoonal northern regions (11–26° S, 129–138° E). Methods With reference to a substantial sub‐continental floristics dataset, we first assigned the fire response (obligate seeder, resprouter) and seedbank persistence (transient, dormant) of rapid and longer‐maturing (> 3 years) woody taxa. Using logistic regression, we then modelled the proportions of taxa possessing these traits as a function of mean annual rainfall (highly correlated with fire frequency) and terrain roughness (a measure of topographic variability) in 0.25° × 0.25° and 1° × 1° grid cells. Separate assessments were undertaken with datasets for 1264 sclerophyll and 236 rain forest taxa. Results This woody flora is characterized by taxa exhibiting mostly resprouting and dormant seedbank traits that promote site persistence. While numbers of obligate seeder and resprouter taxa were related positively to both rainfall and roughness, the relative abundance of both sclerophyll and rain forest obligate seeders decreased significantly with rainfall. The relative abundance of sclerophyll (especially long‐lived) obligate seeders alone increased with topographic roughness. The proportion of taxa with transient seedbanks increased with rainfall in resprouters generally, and in rain forest obligate seeders alone. Main conclusions We find that resprouters are favoured on more productive, fire‐prone sites, and obligate seeders are favoured in less productive, more fire‐protected settings. Seedbank persistence responses are more variable. These findings concur generally with theoretical constructs, and support comparable assessments in Australian and other fire‐prone systems ranging from mediterranean to boreal environments. Our observations illustrate that resprouting and obligate seeding syndromes, but not necessarily seedbank persistence, are useful predictors of vegetation responses to changing fire regime conditions at large landscape scales.     

Allosyncarpia-dominated rain forest in monsoonal northern Australia

Following recent classifications of rain forest vegetation in northern Australia this paper examines the biogeo-graphical status and condition of a rain forest type endemic to that region, dominated by the sclerophyll Allosyncarpia ternata (Myrtaceae). These forests are restricted to the Arnhem Land region of the Northern Territory, which includes Kakadu National Park. They cover an area of 1138 km², or 41 % of all rain forest in northern and northwestern Australia. DCA of floristic data from 140 sites illustrates that Allosyncarpia forests occupy a range of sandstone-derived substrates, from moist valley sediments to steep, freely draining, rocky sites. DCA of floristic transect data illustrates that Allosyncarpia is by far the dominant canopy species over this topographic-moisture sequence, but especially on seasonally dry substrates where it provides over 80 % basal area and effectively the entire canopy. DCA of floristic quadrat data from a floristi-cally singular site illustrates major death of the fire-sensitive gymnosperm Callitris intratropica, and, to a lesser extent, Allosyncarpia itself, on the forest-savanna boundary. Biogeographical implications arising from the Gondwanic distributions of Allosyncarpia and its close relatives, Arillas-trum (New Caledonia), Eucalyptopsis (eastern Malesia), and an as yet undescribed taxon (eastern Australia), suggest that taxa ancestral to this group were extensive in the late Cretaceous. Their current restriction is in marked contrast to the success of their near relatives, the eucalypts. Given the tolerance of Allosyncarpia to a wide range of substrate moisture conditions in the present day, it is argued that fire regulates patch margins of this forest type. Although tolerant of light fires, canopy trees at patch margins are susceptible under a regime of frequent, intense late dry-season fires, such as are prevalent in Arnhem Land today. For effective conservation of fire-sensitive communities in this floristically significant region, greater attention must be given to management and monitoring of the fire regime.     

Australian biogeographical connections and the phylogeny of large genera in the plant family Myrtaceae

Aim To compare the phylogeny of the eucalypt and melaleuca groups with geological events and ages of fossils to discover the time frame of clade divergences. Location Australia, New Caledonia, New Guinea, Indonesian Archipelago. Methods We compare published molecular phylogenies of the eucalypt and melaleuca groups of the plant family Myrtaceae with geological history and known fossil records from the Cretaceous and Cenozoic. Results The Australasian eucalypt group includes seven genera, of which some are relictual rain forest taxa of restricted distribution and others are species‐rich and widespread in drier environments. Based on molecular and morphological data, phylogenetic analyses of the eucalypt group have identified two major clades. The monotypic Arillastrum endemic to New Caledonia is related in one clade to the more species‐rich Angophora, Corymbia and Eucalyptus that dominate the sclerophyll vegetation of Australia. Based on the time of rifting of New Caledonia from eastern Gondwana and the age of fossil eucalypt pollen, we argue that this clade extends back to the Late Cretaceous. The second clade includes three relictual rain forest taxa, with Allosyncarpia from Arnhem Land the sister taxon to Eucalyptopsis of New Guinea and the eastern Indonesian archipelago, and Stockwellia from the Atherton Tableland in north‐east Queensland. As monsoonal, drier conditions evolved in northern Australia, Arnhem Land was isolated from the wet tropics to the east and north during the Oligocene, segregating ancestral rain forest biota. It is argued also that the distribution of species in Eucalyptopsis and Eucalyptus subgenus Symphyomyrtus endemic in areas north of the stable edge of the Australian continent, as far as Sulawesi and the southern Philippines, is related to the geological history of south‐east Asia‐Australasia. Colonization (dispersal) may have been aided by rafting on micro‐continental fragments, by accretion of arc terranes onto New Guinea and by land brought into closer proximity during periods of low sea‐level, from the Late Miocene and Pliocene. The phylogenetic position of the few northern, non‐Australian species of Eucalyptus subgenus Symphyomyrtus suggests rapid radiation in the large Australian sister group(s) during this time frame. A similar pattern, connecting Australia and New Caledonia, is emerging from phylogenetic analysis of the Melaleuca group (Beaufortia suballiance) within Myrtaceae, with Melaleuca being polyphyletic. Main conclusion The eucalypt group is an old lineage extending back to the Late Cretaceous. Differentiation of clades is related to major geological and climatic events, including rifting of New Caledonia from eastern Gondwana, development of monsoonal and drier climates, collision of the northern edge of the Australian craton with island arcs and periods of low sea level. Vicariance events involve dispersal of biota.     

Genetic evidence for the origins of range disjunctions in the Australian dry sclerophyll plant Hardenbergia violacea

Aim The distribution of genetic variation in the Australian dry sclerophyll plant Hardenbergia violacea (Fabaceae) is examined in the context of Pleistocene climate change in order to identify likely refugia. Particular consideration is given to the origin of range disjunctions in South Australia and Tasmania, and to determining whether the Tasmanian population is indigenous or recently introduced from mainland Australia. Location Southeastern Australian mainland and Tasmania. Methods A combination of chloroplast polymerase chain reaction–restriction fragment length polymorphism and genomic amplified fragment length polymorphism (AFLP) marker systems was used to examine the genetic structure of 292 individuals from 13 populations across the range of H. violacea in southeastern Australia. Results Hardenbergia violacea populations in Tasmania and southern Victoria were characterized by low, almost monotypic chloroplast diversity. New South Wales showed higher haplotype diversity and haplotype sharing among widely distributed populations. Principal coordinates analysis (PCoA) of the AFLP data found a strong latitudinal cline in AFLP variation from northern New South Wales south to Tasmania. The Tasmanian population formed an isolated and somewhat disjunct genetic cluster at one end of this cline. However, the South Australian population was an exception to the clinal variation shown by all other populations, forming a highly disjunct cluster in the PCoA. Within‐population genetic diversity was low in both disjunct populations. Main conclusions The genetic evidence indicates that the Tasmanian population is likely to be indigenous and probably the product of vicariance, which was followed by range contraction at the Last Glacial Maximum or an earlier glacial event. The deep phylogenetic disjunction in South Australia is evidence of a much earlier separation on mainland Australia. The chloroplast structure indicates that, during the Pleistocene, H. violacea underwent broad‐scale recolonization in southern Victoria and Tasmania, possibly from a large continental refugium in eastern New South Wales. We conclude that H. violacea, and presumably the sclerophyll communities in which it occurs, have undergone multiple range contractions to large continental refugia during different Pleistocene glaciations in southeastern Australia.     

Development and regional differentiation of the European vegetation during the Tertiary

The Tertiary vegetation of Europe evolved from paratropical to warm-temperate and temperate forms in response to a progressive, non-linear, climatic cooling. Its vegetational forms are composed mainly of two separate ecological units: the evergreen, laurophyll paleotropical geoflora and the deciduous, broad-leaved Arctotertiary geoflora. The development of the Tertiary climate and its interaction with the vegetation are convincingly indicated by the geoflora's migration; the changes in its composition; and the development of the Tertiary forest, swamp, and aquatic plant communities. The paleotropical geoflora is characterized in the upper Cretaceous to the upper Miocene by paratropical rain forest, subtropical rain and laurel forests, temperate laurel forests and edaphically-mediated formation of laurel-conifer forests. The Arctotertiary geoflora advanced into Europe in waves since the Paleocene and formed the basis for the Tertiary mixed mesophytic forests. These can be divided into warm-temperate rain forests, oak-hornbeamchestnut or mixed beech-oak-hornbeam forests, and edaphic formations such as bottomland and swamp forests. Beginning in the lower Cretaceous, the hydrophytic vegetation developed independently of the forest vegetation and formed very diverse herbaceous fresh water, swamp, salt water, and coastal formations. Considerable differences in composition allow to separate floral regions and provinces in Eurosiberia. Instead of three ill-defined floral regions in the Paleocene, there are four well-defined floral regions in the Pliocene. A Mediterranean region cannot be recognized, although Mediterranean (eumesogeic) floral elements appear in the Eocene/Oligocene and thereafter. The Mediterranean sclerophyll forests probably arose after the destruction of the laurophyll forests during the Pleistocene.