Selection of habitat components by two species of agamid lizards in sandridge desert,central Australia

Abstract We studied the use and selection of habitat components by two species of agamid (dragon) lizards in the Simpson Desert, Queensland, Australia. Both the military dragon (Ctenophorus isolepis) and central netted dragon (Ctenophorus nuchalis) were captured in pitfall traps surrounded by areas of open sand, but the mean coverage of spinifex (Triodia basedowii) was greater (35%) around traps capturing C. isolepis than around those capturing its congener (approx. 8%). Direct observations of free‐ranging lizards confirmed that C. isolepis spent most time in, or within 30 cm of, spinifex hummocks, whereas C. nuchalis primarily perched in dead wood. Both species spent most (≥75%) of the time they were observed at rest, with little time spent feeding or moving. Ctenophorus isolepis spent 67% of its time in shade and was active in temperatures between 37.5°C and 43.3°C, whereas C. nuchalis spent 81% of its time in open sun and was active between 35.9°C and 48.1°C. Transect surveys at observation sites showed that open sand was the dominant habitat component available to both species. However, sand was under‐used compared with its availability. Ctenophorus isolepis instead showed strong selection for spinifex cover and C. nuchalis for dead wood; shrub cover was little used. We discuss several mechanisms that may drive the observed pattern of habitat selection, but cannot specify which is the most important. However, the results indicate clear partitioning of habitat between the two species, and suggest that this segregation may facilitate coexistence at local and regional scales.     

Population dynamics of two species of dragon lizards in arid Australia: the effects of rainfall

The population dynamics of two species of agamid (dragon) lizards were studied in the Simpson Desert, central Australia, over a period of 7 years, and modelled in relation to rainfall. Both species have annual life cycles, with adults predominating during the breeding season in spring and summer and juveniles predominating in other seasons. Within years, juvenile abundance in both species in autumn and winter was related most strongly to rainfall in the preceding summer and autumn. This pattern suggests that rainfall enhances survival, growth and possibly clutch size and hatching success. Between years, however, rainfall drove successional change in the dominant plant species in the study area, spinifex Triodia basedowii, causing in turn a shift in the relative abundance of the two species. Thus, the central netted dragon Ctenophorus nuchalis was most numerous in 1990 when vegetation cover was <10%, but declined dramatically in abundance after heavy rainfall at the end of that year. In contrast, the military dragon C. isolepis achieved greatest abundance following heavy rains in the summers of 1990 and 1994, when spinifex cover increased to >20%, and remained numerically dominant for much of the study. We suggest that drought-wet cycles periodically reverse the dominance of the two species of Ctenophorus, and perhaps of other lizard species also, thus enhancing local species diversity over time. Further long-term studies are needed to document the population dynamics of other species, and to identify the factors that influence them. Received: 11 September 1998 / Accepted: 10 February 1999     

Fire and regeneration: the role of seed banks in the dynamics of northern heathlands

Questions: How do species composition and abundance of soil seed bank and standing vegetation vary over the course of a post‐fire succession in northern heathlands? What is the role of seed banks – do they act as a refuge for early successional species or can they simply be seen as a spillover from the extant local vegetation? Location: Coastal Calluna heathlands, Western Norway. Methods: We analysed vegetation and seed bank along a 24‐year post‐fire chronosequence. Patterns in community composition, similarity and abundances were tested using multivariate analyses, Sørensen's index of similarity, vegetation cover (%) and seedling counts. Results: The total diversity of vegetation and seed bank were 60 and 54 vascular plant taxa, respectively, with 39 shared species, resulting in 68% similarity overall. Over 24 years, the heathland community progressed from open newly burned ground via species rich graminoid‐ and herb‐dominated vegetation to mature Calluna heath. Post‐fire succession was not reflected in the seed bank. The 10 most abundant species constituted 98% of the germinated seeds. The most abundant were Calluna vulgaris (49%; 12 018 seeds m^?2) and Erica tetralix (34%; 8 414 seeds m^?2). Calluna showed significantly higher germination the first 2 years following fire. Conclusions: Vegetation species richness, ranging from 23 to 46 species yr^?1, showed a unimodal pattern over the post‐fire succession. In contrast, the seed bank species richness, ranging from 21 to 31 species yr^?1, showed no trend. This suggests that the seed bank act as a refuge; providing a constant source of recruits for species that colonise newly burned areas. The traditional management regime has not depleted or destroyed the seed banks and continued management is needed to ensure sustainability of northern heathlands.     

A simple non-invasive protocol to establish primary cell lines from tail and toe explants for cytogenetic studies in Australian dragon lizards (Squamata: Agamidae)

Primary cell lines were established from cultures of tail and toe clips of five species of Australian dragon lizards: Tympanocryptis pinguicolla, Tympanocryptis sp., Ctenophorus fordi, Amphibolurus norrisi and Pogona vitticeps. The start of exponential cell growth ranged from 1 to 5 weeks. Cultures from all specimens had fibroblastic morphology. Cell lines were propagated continuously up to ten passages, cryopreserved and recovered successfully. We found no reduction in cell viability after short term (<6 months) storage at −80 °C. Mitotic metaphase chromosomes were harvested from these cell lines and used in differential staining, banding and fluorescent in situ hybridisation. Cell lines maintained normal diploidy in all species. This study reports a simple non-invasive method for establishing primary cell lines from Australian dragon lizards without sacrifice. The method is likely to be applicable to a range of species. Such cell lines provide a virtually unlimited source of material for cytogenetic, evolutionary and genomic studies.     

Growth and survivorship as proximate causes of sexual size dimorphism in peninsula dragon lizards Ctenophorus fionni

Males and females differ in body size in many animals, but the direction and extent of this sexual size dimorphism (SSD) varies widely. Males are larger than females in most lizards of the iguanian clade, which includes dragon lizards (Agamidae). I tested whether the male larger pattern of SSD in the peninsula dragon lizard, Ctenophorus fionni, is a result of sexual selection for large male size or relatively higher mortality among females. Data on growth and survivorship were collected from wild lizards during 1991–1994. The likelihood of differential predation between males and females was assessed by exposing pairs of male and female lizards to a predator in captivity, and by comparing the frequency of tail damage in wild‐caught males and females. Male and female C. fionni grew at the same rate, but males grew for longer than females and reached a larger asymptotic size (87 mm vs. 78 mm). Large males were under‐represented in the population because they suffered higher mortality than females. Predation may account for some of this male‐biased mortality. The male‐biased SSD in C. fionni resulted from differences in growth pattern between the sexes. The male‐biased SSD was not the result of proximate factors reducing female body size. Indeed SSD in this species remained male‐biased despite high mortality among large males. SSD in C. fionni is consistent with the ultimate explanation of sexual selection for large body size in males.     

Changes in the stand structure (1975–2000) of coastal Banksia forest in the long absence of fire

Abstract Revisitation studies enable long‐term changes in vegetation to be deciphered and insights into plant community succession to be gained. This is particularly important when assessing the effects of fire exclusion in ecosystems where fire is thought to have once been common. Using two adjacent coastal Banksia integrifolia forest stands in southern Victoria, Australia initially surveyed in 1975 by Hazard and Parsons, we document the changes that occurred in the stand structure between 1975 and 2000. Western Park (WP) has now remained unburnt for over 100 years while Cerberus Naval Base (CNB) was most recently burnt in 1942. Banksia integrifolia densities have decreased at both sites over the 25‐years period by an average of 42–77%, as have other coastal native shrubs (e.g. Leptospermum laevigatum, Leucopogon parviflorus). Trees at WP appear to have died due to old age while mortality at CNB is presumed to be due to stand thinning in response to intense competition for light. Successful recruitment by Banksia has been minimal; trees less than 9 cm girth over bark at breast height (GBBH) were absent at CNB while no trees <19 cm GBBH were observed at WP. The long‐term absence of disturbance such as fire is suspected to be one of the causes of regeneration failure of the stand at WP. Gap phase regeneration is not apparent in B. integrifolia and hence, long‐term succession to a more grassy community is likely when fire is excluded for long periods.     

Why do some species have geographically varying responses to fire history?

A capacity to predict the effects of fire on biota is critical for conservation in fire‐prone regions as it assists managers to anticipate the outcomes of different approaches to fire management. The task is complicated because species’ responses to fire can vary geographically. This poses challenges, both for conceptual understanding of post‐fire succession and fire management. We examine two hypotheses for why species may display geographically varying responses to fire. 1) Species’ post‐fire responses are driven by vegetation structure, but vegetation – fire relationships vary spatially (the ‘dynamic vegetation’ hypothesis). 2) Regional variation in ecological conditions leads species to select different post‐fire ages as habitat (the ‘dynamic habitat’ hypothesis). Our case study uses data on lizards at 280 sites in a ～ 100 000 km² region of south‐eastern Australia. We compared the predictive capacity of models based on 1) habitat associations, with models based on 2) fire history and vegetation type, and 3) fire history alone, for four species of lizards. Habitat association models generally out‐performed fire history models in terms of predictive capacity. For two species, habitat association models provided good discrimination capacity even though the species showed geographically varying post‐fire responses. Our results support the dynamic vegetation hypothesis, that spatial variation in relationships between fire and vegetation structure results in regional variation in fauna–fire relationships. These observations explain how the widely recognised ‘habitat accommodation’ model of animal succession can be conceptually accurate yet predictively weak.     

Microsatellite loci for Australian agamid lizards

We characterize 15 microsatellite loci from two microsatellite libraries developed from the Australian agamid lizards Amphibolurus muricatus and Ctenophorus pictus. All loci were tested for amplification in four other agamids: Ctenophorus fordi, Ctenophorus decresii, Chlamydosaurus kingii, and Physignathus lesueurii. These loci were highly polymorphic within and across species, with nine to 12 loci amplifying for each species tested.     

Fire history and vegetation recovery in two raised bogs at the Baltic Sea

Questions: What were the bog fire patterns and frequencies in two boreal peatlands during the last 5000 years? What is the nature and time‐scale of post‐fire vegetation successions? Were fire events related to climate? Location: Männikjärve bog, central east Estonia; Kontolanrahka bog, southwest Finland. Methods: Macroscopic charcoal, plant macrofossils and radiocarbon dating were examined. Redundancy analysis was used in the assessments. Results: During the last 5000 years, both of the above peatlands have experienced several fire events. A typical pre‐fire vegetation community consisted of dry hummock Sphagnum spp., often accompanied by Calluna vulgaris. Only the most severe occasional fires resulted in a dramatic change in the vegetation composition. In these cases, a wet shift occurred, where the pre‐fire hummock community was replaced by a wet hollow community. Calluna vulgaris was found to be a key species in both pre‐ and post‐fire vegetation dynamics. The recovery time of dry microtopes following severe combustion and the subsequent hydrological change could take up to 350 years. Even after a long‐lasting wet phase, the post‐fire disturbance succession led towards a dry hummock community. Conclusions: Fire succession appeared to be cyclic, starting as and developing towards a dry hummock community. Fires have been a regular phenomenon in boreal bogs, even in regions with rather low human impact. The fire history records did not indicate any direct link to the regional long‐term climate.     

Influence of fire history on small mammal distributions: insights from a 100‐year post‐fire chronosequence

Aim Fire affects the structure and dynamics of ecosystems world‐wide, over long time periods (decades and centuries) and at large spatial scales (landscapes and regions). A pressing challenge for ecologists is to develop models that explain and predict faunal responses to fire at broad temporal and spatial scales. We used a 105‐year post‐fire chronosequence to investigate small mammal responses to fire across an extensive area of ‘tree mallee’ (i.e. vegetation characterized by small multi‐stemmed eucalypts). Location The Murray Mallee region (104,000 km²) of semi‐arid Australia. Methods First, we surveyed small mammals at 260 sites and explored the fire responses of four species using nonlinear regression models. Second, we assessed the predictive accuracy of models using cross‐validation and by testing with independent data. Third, we examined our results in relation to an influential model of animal succession, the habitat accommodation model. Results Two of four study species showed a clear response to fire history. The distribution of the Mallee Ningaui Ningaui yvonneae, a carnivorous marsupial, was strongly associated with mature vegetation characterized by its cover of hummock grass. The occurrence of breeding females was predicted to increase up to 40–105 years post‐fire, highlighting the extensive time periods over which small mammal populations may be affected by fire. Evaluation of models for N. yvonneae demonstrated that accurate predictions of species occurrence can be made from fire history and vegetation data, across large geographical areas. The introduced House Mouse Mus domesticus was the only species positively associated with recently burnt vegetation. Main conclusions Understanding the impact of fire over long time periods will benefit ecological and conservation management. In this example, tracts of long‐unburnt mallee vegetation were identified as important habitat for a fire‐sensitive native mammal. Small mammal responses to fire can be predicted accurately at broad spatial scales; however, a conceptual model of post‐fire change in community structure developed in temperate Australia is not, on its own, sufficient for small mammals in semi‐arid systems.     

Effects of landslides on the mountain vegetation of Flores Island,Azores

Question: What are the consequences of frequently occurring landslides on vegetation dynamics, floristic and structural diversity? Location: 39°27′N; 31°13′W – Morro Alto, Flores Island, Azores, Portugal. Methods: Six comparable landslides were selected. Plots were placed at the top, slope and toe of landslides. Data on floristic composition and biovolume, demography and size structure of the dominant tree species (Juniperus brevifolia) were collected. Hierarchical agglomerative clustering and Principal Component Analysis were used in order to identify succession stages and compare succession pathways and vegetation recovery in different parts of the landslides. Results: Four stages of primary succession on substrates formed by landslides were identified: pioneer (Festuca‐Sphagnum grassland), assembly (Juniperus‐Festuca‐Sphagnum open scrub), building (Juniperus‐Sphagnum scrub) and mature (Juniperus‐Sphagnum woodland). Concerning J. brevifolia populations, the succession pathways are independent of location on the landslide. However, at the floristic level, there are some differences, mainly in the pioneer stage at the toes of landslides. Better abiotic conditions, resulting in a higher succession rate, are probably responsible for a faster vegetation recovery on landslide toes. Conclusion: Landslides trigger succession processes that enable a massive regeneration of the dominant tree species and existence of species not present in mature forests. They are also responsible for the simultaneous occurrence of vegetation of different structures. Overall, landslides increase the floristic and structural diversity of the vegetation, consequently increasing landscape heterogeneity.     

Fire and succession in the ultramafic maquis of New Caledonia

Aim This study investigates the role of fire and post fire succession in determining the structure and composition of vegetation on ultramafic iron crust soils. Location The study was conducted in the Plaines des Lacs region of southern New Caledonia. Methods A survey was made of eighty-eight sites, recording floristic composition, trunk size-class distributions, regeneration after fire, growth ring counts of Dacrydium araucarioides (Podocarpaceae) and historical information on past fires. Floristic data was ordinated using multidimensional scaling and an index of succession based on structural and historical information. A transition matrix model was developed to predict the effect of fire frequency on vegetation composition. Results The vegetation is undergoing postfire succession from maquis to forest, after about 75 years, and eventually to rainforest. Gymnostoma deplancheanum has a key role as an early colonist that produces shade, the bulk of the litter, and forms nitrogen fixing nodules with Frankia sp. However, the open canopy of Gymnostoma and slow litter decay creates flammable conditions. Though many species resprout from rootstocks, only thirty-nine persist through fires while 114 others colonize at later successional stages, as the litter layer and shade increase. Some early successional species are later excluded but these can persist locally in swamps and on rocky hill tops. Forest and rainforest are less flammable and the matrix model suggests that ignition frequency has a critical role in determining the abundance of maquis or forest. Main conclusions The vegetation mosaic represents a post fire succession from open maquis to forest. Palynological and charcoal records from late Pleistocene sediments suggest that fire has been a major factor determining the development of maquis vegetation since before the arrival of humans. Recently, frequent fires have converted much of the vegetation to maquis, posing a threat to some forest species and largely eliminating rainforest from iron crust soils.     

Correlates of active body temperatures and microhabitat occupation in nine species of central Australian agamid lizards

Body temperatures of active lizards and their correspondence with microhabitat occupation were studied for nine species of agamid lizards in the central Australian arid zone. Thermoregulatory behaviour was also documented using several measures, such as the use of shade and perch height. The effects of thermal environment on lizard habitat occupation were hypothesized to be significant, because desert regions experience daily and seasonal extremes of temperature that are well in excess of a lizard’s preferred temperature range. All species, except Ctenophorus isolepis and Diporiphora winneckei, were found to have body temperatures that corresponded closely to ground and surface temperatures. Thermoregulatory behaviour was also found to be important throughout a lizard’s daily activity; all study species, other than Ctenophorus isolepis, were found to increase their perch height in the middle of the day. Ctenophorus isolepis was shown to be a strictly terrestrial species that uses the shade of spinifex in its thermoregulatory behaviour. Species exhibited a non‐random selection of microhabitats and a preference for a particular set of thermal and structural factors. In this study, it was shown that structural factors were particularly important in microhabitat occupation. Thermal factors accounted for a smaller proportion of variance in microhabitat occupation, but still played a considerable role in the microhabitat use in central Australian agamids.     

Responses of two species of heathland rodents to habitat manipulation: Vegetation density thresholds and the habitat accommodation model

The abundance of two native rodent species, Rattus lutreolus and Pseudomys gracilicaudatus, has been shown to correlate with vegetation density in coastal wet heath. Fox's habitat accommodation model relates relative abundances of such small mammal species to heathland vegetation regeneration following disturbance. Implicit in the model is recognition that it is successional changes in vegetation, not time per se, that drives the responses of small mammal species along a regeneration axis. Using a brush‐cutter we deliberately removed approximately 85% of vegetation around trapping stations and recorded significant reductions in the abundance of both P. gracilicaudatus (an earlier‐stage colonizing species) and R. lutreolus (a late seral‐stage species). A significant decrease in the abundance of only the latter had been demonstrated previously when 60–70% of the vegetation had been removed. Following the brush‐cutting both species re‐entered the mammalian secondary succession at different times, first P. gracilicaudatus followed by R. lutreolus after the vegetation cover thresholds of each species had been reached. The impact of this habitat manipulation experiment was to produce a retrogression of the small mammal succession, experimentally demonstrating causality between changes in vegetation density and subsequent small mammal habitat use.     

Contrasting changes in vegetation structure and diversity with time since fire in two Australian Mediterranean‐climate plant communities

Mallee‐heath and mallee communities occur in a mosaic across large areas of south‐western Australia, in topographically subdued and fire‐prone landscapes. Consequently, it could be expected that these communities would have historically experienced similar fire regimes, and would respond similarly to variation in aspects of the fire regime. We studied the response of mallee‐heath and mallee to time since the last fire, measuring species density, species–area relationships, diversity indices and vegetation structure. Floristic responses to time since fire accorded with the initial floristic composition model of plant succession, with declining species density and Shannon diversity with age in mallee‐heath. Mallee‐heath exhibited structural senescence when > approximately 45–55 years since fire, with increasing standing dead vegetation, bare ground and stagnating or declining size in sprouting Eucalyptus spp. Mallee showed no such evidence of senescence, and indeed continued to increase in stature beyond the mean fire interval but without the compositional change required to provide support for the relay floristic model of plant succession. These results indicate that mallee‐heath is a fire maintained community and as such is reliant upon periodic burning to maintain diversity and vigour. Mallee, in contrast, is modified but not maintained by fire (at least over the period of time since fire examined) and hence is less susceptible to fire interval effects. Indeed, structural attributes likely to be significant for fauna habitat and carbon sequestration continue to develop in mallee unburnt for 55 years or more. Different responses to time since fire will create challenges for management, particularly in fragmented landscapes where fire potentially interacts with other threatening processes.     

Polymorphic microsatellite markers in the painted dragon lizard,Ctenophorus pictus

We isolated eight polymorphic microsatellite markers from the painted dragon, Ctenophorus pictus, from a genomic library simultaneously enriched for a suite of di‐, tri‐ and tetranucleotide repeat units. Genetic variation was assessed in 40 individuals from populations in Victoria and South Australia. The number of alleles per locus ranged from four to 45 with observed heterozygosity ranging from 0.23 to 0.95. These markers will be useful for studies of natural and sexual selection and the effects of habitat fragmentation in semiarid Australia.     

Contrasting effects of landscape features on genetic structure in different geographic regions in the ornate dragon lizard,Ctenophorus ornatus

Habitat fragmentation can have profound effects on the distribution of genetic variation within and between populations. Previously, we showed that in the ornate dragon lizard, Ctenophorus ornatus, lizards residing on outcrops that are separated by cleared agricultural land are significantly more isolated and hold less genetic variation than lizards residing on neighbouring outcrops connected by undisturbed native vegetation. Here, we extend the fine‐scale study to examine the pattern of genetic variation and population structure across the species' range. Using a landscape genetics approach, we test whether land clearing for agricultural purposes has affected the population structure of the ornate dragon lizard. We found significant genetic differentiation between outcrop populations (F_ST = 0.12), as well as isolation by distance within each geographic region. In support of our previous study, land clearing was associated with higher genetic divergences between outcrops and lower genetic variation within outcrops, but only in the region that had been exposed to intense agriculture for the longest period of time. No other landscape features influenced population structure in any geographic region. These results show that the effects of landscape features can vary across species' ranges and suggest there may be a temporal lag in response to contemporary changes in land use. These findings therefore highlight the need for caution when assessing the impact of contemporary land use practices on genetic variation and population structure.     

Effect of legume species introduction to early abandoned field on vegetation development

One of the most important areas in ecology is to elucidate the factors that drive succession in ecosystems. The purpose of our study was to assess the effects of legume species (Medicago sativa, Melilotus suaveolens and Astragalus adsurgens) introduction to abandoned arable land on vegetation development in the Loess Plateau, China. Results from our study showed that addition of legume species strongly affected the composition of recently abandoned-field vegetation. Legume species were effective at reducing the number and dominance of natural colonizers (mainly weeds from the seed bank). The introduction of legume species into newly abandoned fields maintained high total cover and above-ground biomass and could improve soil organic carbon and total nitrogen. However, the effects of the treatments were species-specific. Melilotus suaveolens turned out to be severely suppressive to natural colonizers (weed species). Also, Melilotus suaveolens-adding maintained the highest cover and above-ground biomass and was helpful to improve later succession species, e.g. Stipa breviflora and Astragalus polycladus, to invade and establish. Medicago sativa-adding was superior in enhancing the soil organic carbon and total nitrogen. The present results suggested that addition of legume species with greater cover and biomass strongly suppressed the dominance of the weedy species in early succession and the course of old-field succession may be accelerated by introduction of legume species at least temporarily. However, the experimental period was too short to assess to what extent succession may be affected in the longer term.     

The age of Calluna stands moderates post‐fire regeneration rate and trends in northern Calluna heathlands

Questions: Does stand age influence the direction and rate of post‐fire successional dynamics in coastal Calluna heaths and can old degraded heath vegetation be restored through reintroduction of fire? Location: Coastal heaths in the Tarva archipelago, central Norway. Methods: We investigated revegetation dynamics after experimental fires set in young (8 years since last fire) and old (>50 years since last fire) grazed heath stands. A repeated measures design was used, with floristic data recorded in permanent plots in the post‐fire successions (n=12) over a 7‐year period. The data were analysed using multivariate ordination techniques (PCA, RDA and PRC) and mixed effects models. Results: The age of Calluna stands strongly influenced post‐fire succession, different trends due to age explained 10.4% of variation in floristic data. Young heath showed faster succession towards pre‐fire community composition than old heath, and this could partially be explained by succession‐related factors: young heath had lower cover of mosses and lichens in the pre‐burned vegetation, and lower cover of litter early in succession. Young heath had a less pronounced overall community response to fire than old heath. Vegetative regeneration of C. vulgaris was absent in both old and young heath, but Calluna still re‐established as the dominant species within 5–7 years in both young and old stands. Regeneration dynamics were also affected by habitat conditions, different trends due to habitat explained 6% of variation. Conclusions: Our study demonstrates that old stands do develop characteristic heathland vegetation and structure after fire, and while potential invasives into the system such as trees and rhizomatous species are present, they do not impair Calluna regeneration or vegetation development towards the target heathland community composition and structure. Further, as our young stands are only in their second fire rotation after restoration, we suggest that characteristic dynamics of managed heathlands can re‐establish relatively rapidly, even in severely degenerated sites (>50 years since last fire). Site‐specific factors also need to be considered. We conclude that there is restoration potential in old heaths, despite slow dynamics in the first rotation.     

Recovery from clearing,cyclone and fire in rain forests of Tonga,South Pacific: Vegetation dynamics 1995–2005

Abstract In late 2001 a category 3 cyclone impacted forest plots that were established in Tonga in 1995, and additionally, one plot was accidentally burned by an escaped land‐clearing fire. Subsequent surveys provide observations of 10 years of forest dynamics in this poorly studied region, and the first reported observations of large interannual variation in juvenile (seedling and sapling) abundance in the western tropical Pacific. The severely disturbed (burned) plot was initially colonized by a non‐native early pioneer, Carica papaya L., but 3.5 years later a native pioneer, Macaranga harveyana (Muell. Arg.) Muell. Arg., was the most abundant tree species. The seedling layer included some long‐lived pioneers and shade‐tolerant species. Two mature forest plots affected only by the cyclone changed very little over a decade. Late‐successional shade‐tolerant species that dominated the overstory were also abundant as seedlings and saplings. This is in contrast with a 30‐ to 40‐year‐old, formerly cultivated, secondary forest plot that still shows no recruitment of late‐successional dominants, in spite of the proximity of remnant forest patches. This study suggests differing pathways of succession following shifting cultivation versus cyclone and fire disturbances in Tonga. Land use legacies appear to have a long‐lasting effect on community composition.