Figures
Abstract
A variety of environmental processes, including topography, edaphic and disturbance factors can influence vegetation composition. The relative influence of these patterns has been known to vary with scale, however, few studies have focused on environmental drivers of composition at the mesoscale. This study examined the relative importance of topography, catchment flow and soil in influencing tree assemblages in Karawatha Forest Park; a South-East Queensland subtropical eucalypt forest embedded in an urban matrix that is part of the Terrestrial Ecosystem Research Network South-East Queensland Peri-urban SuperSite. Thirty-three LTER plots were surveyed at the mesoscale (909 ha), where all woody stems ≥1.3 m high rooted within plots were sampled. Vegetation was divided into three cohorts: small (≥1–10 cm DBH), intermediate (≥10–30 cm DBH), and large (≥30 cm DBH). Plot slope, aspect, elevation, catchment area and location and soil chemistry and structure were also measured. Ordinations and smooth surface modelling were used to determine drivers of vegetation assemblage in each cohort. Vegetation composition was highly variable among plots at the mesoscale (plots systematically placed at 500 m intervals). Elevation was strongly related to woody vegetation composition across all cohorts (R2: 0.69–0.75). Other topographic variables that explained a substantial amount of variation in composition were catchment area (R2: 0.43–0.45) and slope (R2: 0.23–0.61). Soil chemistry (R2: 0.09–0.75) was also associated with woody vegetation composition. While species composition differed substantially between cohorts, the environmental variables explaining composition did not. These results demonstrate the overriding importance of elevation and other topographic features in discriminating tree assemblage patterns irrespective of tree size. The importance of soil characteristics to tree assemblages was also influenced by topography, where ridge top sites were typically drier and had lower soil nutrient levels than riparian areas.
Citation: Hero J-M, Butler SA, Lollback GW, Castley JG (2014) Determinants of Tree Assemblage Composition at the Mesoscale within a Subtropical Eucalypt Forest. PLoS ONE 9(12): e114994. https://doi.org/10.1371/journal.pone.0114994
Editor: Paul Adam, University of New South Wales, Australia
Received: May 20, 2014; Accepted: November 17, 2014; Published: December 12, 2014
Copyright: © 2014 Hero et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. Data are available at figshare.com (DOI: http://dx.doi.org/10.6084/m9.figshare.1221851).
Funding: Funded by the Brisbane City Council, Griffith University, SEQ Catchments and TERN Australian SuperSite Network. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
A number of environmental processes influence vegetation patterns, which is typically observed as variation in vegetation structure and composition as a continuum along environmental gradients [1]. The relative influence of environmental processes however, varies at different spatial and temporal scales [2], [3], [4]. For example, it is widely accepted that regional vegetation patterns (10–1000 km2) are determined by regional variation in climate, landform and soil type [2], [5], [6], [7], [8], [9], [10], whereas local variation (<1 km2) is determined by finer scales of soil chemistry, disturbance and biological interactions [11], [12], [13], [14]. What is less clear is how the role of these processes changes from local to regional scales due to few studies being undertaken at mesoscales (a study area of 1–10 km2 size with adequate coverage of replicates, [16], [17]).
The relative importance of environmental factors in influencing vegetation patterns has had some attention, with most research focusing on the associations of individual or functional groups of factors such as soil chemistry [18] or topography [19], [20]. These studies typically demonstrate highly significant correlations between species patterns and environmental factors, although correlations at the assemblage level are often relatively low. The few studies that have evaluated a range of factors have demonstrated that individual factors alone do not explain the observed spatio-temporal patterns of vegetation [16]. Consequently vegetation patterns within a landscape may therefore be driven by a number of environmental processes.
The importance of environmental factors on eucalypt community and species levels patterns has also been widely documented, although most studies have occurred at regional [21], [22], [23], [24], [25] or local scales [12], [26], [27]. For example, variations in eucalypt savannas at the regional scale are attributable to soil type gradients [21], [22], [23] or very broad elevation gradients [28], [29]. Similarly, local scale variation in subtropical eucalypt forests varies significantly between valley-floors and ridge-tops [27], which is attributable to hydrological and soil nutrient gradients within a catchment.
The strong associations between vegetation patterns and the environment may also result in structural and compositional differences in size cohorts within an assemblage. This is evident in the marked differences in the determinants of the potential and observed distribution and abundance patterns of seedlings and mature sized vegetation [30]. For example, the potential distribution and abundance of seedlings is codetermined by seed dispersal patterns and seed output, which are in turn influenced by environmental factors and biological interactions, as well as genotype specific effects of individual seeds [31]. In contrast, determinants of realised abundance patterns of seedlings and subsequent abundance within size classes are influenced by a suite of environmental conditions [30], [31]. The differences between the environmental determinants of trees of different size cohorts however have been little studied, with most studies focusing on diameter density curves and predicting population trends [30], [32], [33], [34] rather than on determinants of recruitment variability between size cohorts at the assemblage level.
This study evaluates the role of a range of environmental factors in influencing the variation in tree assemblages across a complex subtropical eucalypt forest mosaic at the mesoscale in eastern Australia. At the mesoscale, the importance of a number of habitat variables including topography, available runoff and soil chemistry and structure may be responsible for determining vegetation patterns among size cohorts. This will be tested independently for trees within three size cohorts at the assemblage level. The study of these relationships will provide new insight into the relative importance of environmental determinants of assemblages at a scale and in a forest type that has received little attention in the scientific literature.
Materials and Methods
Site Description
This study was conducted at Karawatha Forest Park (KFP) in southeast Queensland, Australia (Fig. 1), which is located 18 km south of Brisbane (Latitude −27.632 and Longitude 153.084). It is a long-term ecological research site (LTER node) that is part of the South-east Queensland Peri-urban Supersite within the Terrestrial Ecosystem Research Network (TERN). More information about TERN Supersites can be found at http://www.tern-supersites.net.au. The forest park covers an area of approximately 900 ha and is characterised by remnant mixed eucalypt forest with a predominately grassy understorey and melaleuca forest with a fern and herbaceous understorey. It is one of the largest remaining peri-urban forests under governance of the Brisbane City Council, with one of the highest richness of eucalypt species within the greater Brisbane region [35]. The vegetation communities are divided into two major groups including wetlands and mosaic mixed eucalypt communities, with the wetlands restricted to areas along the creek systems and dominated by Melaleuca quinquenervia. In contrast, the well-drained and infertile sandstone ridgelines in the north of Karawatha are dominated by eucalypt communities [35].
The midlines (thick black lines), and 10 m contour lines (thin grey lines with values in metres) demonstrate the positions for thirty three PPBio LTER plots surveyed in this study. The inset shows the location of KFP (star) within Queensland, Australia. Each midline starting point was placed systematically on a grid. These grid locations are also displayed as a combination of a letter and number.
Experimental Design
Fieldwork was conducted between February and July 2007. Permits were not needed for this study and permission to perform vegetation and soil sampling within KFP was granted by Brisbane City Council (BCC), which governs KFP. The BCC is the single local government authority responsible for the administration of the whole of the metropolitan area of Brisbane. No destructive sampling was undertaken during this study. Sampling at 33 permanent plots followed the Planned Program for Biodiversity Studies (PPBio) protocols developed by Magnusson et al. [36] and described by Hero et al. [37] and Magnusson et al. [38]. The plot locations were positioned using a systematic grid system with dimensions of approximately 3.5 km by 3 km comprised of 8 east-west and 6 north-south gridlines. Grid placement was randomly determined by desktop mapping to maximise the number of plots within the park boundaries. The positioning of plots between gridlines maximised the distance between plots, such that each plot begins approximately 500 m from neighbouring plots. Plot coordinates were determined from desktop mapping and were located on site using a GPS (Fig. 1).
Each plot is 250 m long and 42 m wide with a 2 m wide midline dividing the plot into two 20 m wide sections, with a total sample area of 1 ha. Each 1-ha plot follows the topographic contours thereby minimising within plot variation and maximising between plot variation in terms of catchment area, soil types, topography, and vegetation structure and composition facilitating robust comparison between plots [16], [37], [39]. The plots were permanently established using surveying equipment with metal pegs spaced along the plot midline at 10 m intervals to allow for long-term research and ecological monitoring as part of PPBio in Australia.
A hierarchical spatial design was used to sample woody vegetation ≥1.3 m in height from different size cohorts: trees ≥1–10 cm DBH were sampled within 0.1 ha (4 m×250 m); trees ≥10–30 cm DBH were sampled within 0.5 ha (20 m×250 m), and woody vegetation ≥30 cm DBH were sampled within 1 ha (40 m×250 m). The size classes used in this study are designed to adequately estimate the density of woody species of different sizes that vary in their horizontal spatial density (sensu Costa et al. [16]). For example, saplings often present at higher densities than larger trees due to their relatively higher recruitment rates [30], [32] and therefore require a relatively smaller area to estimate densities.
The hierarchical sampling design was used to tag and measure the DBH of all trees rooted within plots. Individuals within Myrtaceae, Proteaceae, Fabaceae, Rhamnaceae and Casuarinaceae were identified to species level where possible, including those from the genera Acacia, Allocasuarina, Alphitonia, Angophora, Banksia, Corymbia, Callistemon, Eucalyptus, Melaleuca, Lophostemon and Persoonia using taxonomic classification by the Queensland Herbarium [40] for eucalypts and Flora of Australia [41], the Australian Biological Resources Study [42] and the Australian Biological Resources Study [43] for all other species. The abundance of species was collated within each size class for each plot.
The environmental variables measured at each plot included plot topography, soil texture and chemistry. Topographical measurements included altitude, slope, aspect and available runoff was represented by catchment area. Plot altitude was measured from topographical maps, aspect was measured using a compass, and slope was measured using clinometers over a distance of 2 m in the direction of the greatest slope with measures taken every 50 m and averaged for each plot. Catchment area was measured using topographic maps, with catchment area digitized using Autocad 2006 [44] and measured in square metres.
Soil samples within each plot were taken to a depth of 10 cm below the surface of soil at six points (0, 50, 100, 150, 200 and 250 m) along the plot midline and assembled to produce a composite sample for each plot. Prior to analysis, the composite soil samples were dried at room temperature, cleaned of stones and fine roots, and passed through a 2 mm sieve according to the protocol outlined by Kalra and Maynard [45]. Soil particle size distribution (% silt, clay and sand) was determined using the hydrometer method (sensu Kalra and Maynard [45]). Soil chemistry analyses included pH, Electrical Conductivity (EC), total carbon (C), total nitrogen (N), total phosphorus (P), Cation Exchange Capacity (CEC), magnesium (Mg2+), potassium (K+), sodium (Na+), calcium (Ca2+), nitrates (NO3−) and nitrites (NO2+) using standard methods described by Rayment and Higginson [46].
Analyses
Trees were split into the size cohorts of 1–10 cm DBH, 10–30 cm DBH and>30 cm DBH for all analyses. This is because different selection pressures may affect each size class [31]. All analyses involved a measure of ecological resemblance and we chose to use Gower's similarity coefficient (equation 7.20 in Legendre and Legendre [47]), because it is symmetrical and accounts for the ecological significance of the absence of a species in two or more sites.
The spatial arrangement and autocorrelation of species composition was evaluated using a permutated (n = 999) Mantel statistic. The statistic was calculated for distance classes that reflected the spatial arrangement of PPBio plots. A significance test (P<0.05) indicates either spatial autocorrelation with positive Mantel statistic values, or dispersion with negative statistic values [47].
The relationship between tree species composition and the environment was examined using indirect gradient analysis. If two explanatory variables had a correlation coefficient>0.7, one of the variables was not included in the analysis [48]. Because of the many soil variables (n = 14), a Principal Components Analysis was used to produce major axes from the soil data. The major axes scores were then used as predictor variables. Species composition data were first analysed using a non-metric multidimensional scaling analysis (NMDS). The number of axes used was decided by plotting the stress value against the number of axes. The number of axes was chosen when the reduction of stress stopped decreasing noticeably with increasing axes. The data were centred, rotated and scaled for the NMDS analysis. Smooth surfaces of the explanatory variables were then fitted to the NMDS ordination. The strength of the relationship between the ordination and the fitted environmental vectors is reflected in an R2 coefficient of determination. All analyses were carried out in the R program [49] using the package vegan [50].
Results
Overall, 49 woody species were identified from 10 825 sampled individuals. The most common genus was Eucalyptus, which made up 33% of the coverall composition, while Lophostemon, Corymbia, Melaleuca and Acacia were also well represented genera (Table 1). Species composition differed between cohorts, with species diversity being the most even within the youngest cohort (Table 1, Figs. 2 and 3). As the DBH of the cohorts increased, the proportion of trees that were eucalypts also increased.
Although three NMDS axes were used, only the first two are shown for visual representation.
Spatial correlation
There was no significant spatial autocorrelation of woody vegetation composition over a distance 500 m to 2000 m (rM 0.08–−0.09). However, there was significant dispersion of trees with a DBH of 10–30 cm and>30 cm DBH at a distance of 2000 m among plots (rM = −0.08 and −0.09, respectively. P = 0.03 and 0.02, respectively).
Patterns of Vegetation Assemblages along Environmental Gradients
Three-dimensional NMDS with a stress of 0.18, 0.15 and 0.15 (small to largest cohorts, respectively) represented the input matrix of vegetation composition and described similarity between plots for all cohorts (Fig. 3). Three soil PCA axis were used for the surface analysis, with Axis I, II and III explaining 45%, 13% and 12% of the variation of the soil correlation matrix, respectively. Axis I was negatively related to Mg2+, Na+, EC, CEC, C and N (Table 2). The second component was positively related to Ca2+ and P and negatively related to K+ and Na+. Axis III was highly positively associated with NO3−, positively related to NO2− and negatively related to the proportion of silt and pH.
Surface modelling indicated that elevation, catchment area, and slope significantly influenced species composition of all cohorts (Table 3). All of these variables can be classed as topographic variables. Elevation explained the most variation in species composition among all cohorts, as demonstrated by the changing composition and relative abundance of species along the elevation gradient (Fig. 4). Soil Axis I was associated with composition in the smallest and largest cohorts (Table 3).
Discussion
This study has shown that variation in tree assemblages in a subtropical eucalypt forest at the mesoscale is strongly related to topographic variables and secondarily related to soil properties. Elevation clearly explained the most variation of composition in all cohorts. The composition of size cohorts was different; however, the strength of environmental variables explaining composition only differed moderately. Composition of all cohorts was not spatially correlated between plots.
Patterns of Vegetation Assemblages along Environmental Gradients
Elevation has been shown to drive vegetation patterns including functional traits, species richness, species composition and density in numerous studies at a number of spatial scales [51], [52], [53], [54], [55], [56], [57], [58] and in a number of vegetation types [27], [59]. Elevation as an environmental correlate however, has no direct physiological impact on vegetation patterns [60], [61] and acts as a surrogate for the influence of direct factors including catchment flows, slope and soil chemistry (Table 4). A range of environmental variables including hydrology and soil chemistry have also been shown by other studies to vary along narrow elevation gradients [16], [20], [60], [61], [62], [63]. The strong association between a narrow elevation gradient and vegetation patterns as found in this study reflect the role of numerous direct environmental factors [61].
Catchment area can influence vegetation patterns because small catchment area represents a relatively smaller amount of available water compared to large catchment areas, especially at the lower areas of the catchment. The influence of water availability including catchment runoff [64] and regional hydrological patterns [64], [65] on vegetation assemblage development has been shown to be closely linked to species habitat requirements. Species with similar hydrological requirements are consequently clustered in optimal areas within catchment systems, resulting in specific assemblage patterns.
Like elevation, slope can indirectly influence habitat suitability and thereby composition at local scales [20], [66] by changing microclimate, runoff, drainage, soil erosion and formation [57], [67]. Although not always the case [16], it appears slope can be a strong driver of species composition that can influence over small [57], [68], meso [15] and large scales [67], [69]. Slope was an important driver in KFP for all cohorts, but more so for larger cohorts.
Although topography influences soil characteristics and thereby drives tree composition, soil chemistry, as measured in this study, generally did not describe tree composition variability as much as topography at KFP. This reflects the findings of Catterall et al. [27], who found that elevation, not measured soil properties, was important in explaining eucalypt composition in a nearby forest patch to KFP. In contrast, Wardell-Johnson et al. [70] showed that soil chemistry was associated with vegetation composition in the same study area used by Catterall et al. [27]. Different sampling regimes were used between the two studies, with coverage and the number of replicates of Wardell-Johnson et al. [70] being comparatively higher. On a broad scale, Wills and Clarke [71] summarised that Australian vegetation types change with rainfall, temperature and soil. Further afield, Pansonato et al. [72] summarised that many studies in western Amazon have found that soil fertility was an important driver of plant communities, while central Amazonian studies have found that topography or soil texture were predictors of vegetation communities. Further studies have also found that topography and soil influence vegetation composition at small [68] and large scales [4] and Eiserhardt et al. [3] suggested that climate was more of a driver of palm vegetation patterns at broader scales, but topography and soil influenced vegetation at smaller scales. Considering there is some, but not complete correlation between soil and topography, it is not surprising that both soil chemistry and topography influenced different cohorts within KFP at the mesoscale. Pansonato et al. [72] used PPBio plots in Brasil to show that it was the amplitude of environmental factors and not the scale of these factors that was important in explaining composition. While the elevation only ranged from 28–80 m at KFP, the soil types did vary from well-drained, infertile sandstone on the ridges to richer alluvial soil in the wetlands. Similarly, this variation was also associated with the turnover of the woody vegetation community at KFP (Fig. 4).
The gradient of vegetation from ridge tops to valley floors detected here (Fig. 4) is consistent with similar studies on a variety of vegetation types and spatial scales [17], [20], [27], [55]. This model is characterised by the inherent differences between valley floors and ridge tops, where valley floors have relatively higher soil moisture, a shallower water table, and higher nutrient loading compared to upslope and ridge top areas [17], [20], [27]. These studies however, did not sample over the entire catchment gradient, instead only sampling at the tops of ridges and at the bottom of valley floors which provides an easily detected difference. Our study therefore contributes to an important, although generally only assumed, body of literature on the gradient of tree assemblages driven by water and nutrient availability.
The eucalypt assemblages at Karawatha were distributed along a gradient of changing tree species abundance without clear boundaries. This spatial pattern is in contrast to the community-unit theory where plant communities form homogenous, distinct and identifiable units [1]. However, the inherent difference between observing vegetation patterns as communities or as changing assemblages along a continuum depends on the spatial and temporal scale of the study [12], [73]. For example, Battaglia and Williams [12] demonstrated that eucalypt savannas changed along a continuum if viewed at the soil nutrient scale, although clear community boundaries were present if viewed at the soil type scale. This demonstrates that at the mesoscale, catchment flow and soil gradients may be too narrow to produced identifiable vegetation communities due to the lack of distinct environmental habitats along the gradient from ridge-tops to riparian areas, although large enough to detect significant change in species assemblages across the catchment gradient.
Differences between Vegetation Size Cohorts
While elevation was one of the best environmental determinants of tree assemblages, there were slight variations in the relative importance of other environmental variables for the three size cohorts (Table 4). Slope explained more variation in the two larger cohorts, soil was not a significant predictor in the middle cohort and catchment area was a similarly important driver within all cohorts. While these differences may reflect the role of selective pressures on the abundance of trees during different stages of their life history [31] the differences detected in our models are relatively small. This suggests that the selective pressures of species abundance patterns in subtropical eucalypt forest may be relatively constant through each life cycle phase.
Despite little difference in the environmental determinants there were substantial differences in species composition and abundance within the three size cohorts (Figs. 2 and 3). The dissimilarity between size cohorts appears to be related to differences in the life history strategies of species dominating each size cohort. Pioneer species and species restricted to the understorey such as Lophostemon spp., Acacia spp. and Allocasuarina littoralis were relatively common in the smaller cohort, but became less common as cohort size increased. Intermediate and large trees from the genera Eucalyptus and Corymbia dominated the larger size cohorts. These differences reflect the turnover or succession of species through each consecutive size cohort [74]. Fire may also shape the vegetation composition and structure by killing a disproportionate number of trees in the smallest cohort [75]. The structural and compositional differences between size cohorts therefore reflect the interaction between tree species life history strategies and environmental processes.
Future Directions and Management Implications
Studying plant physiology at KFP would likely shed more light on the mechanisms behind species turnover. This is because certain physiological characteristics such as leaf mass, stem height and density, root depth and seed size are likely to favour certain environmental conditions and therefore differ between species [76], including between eucalypt species [77]. Less than 300 km away from KFP, Kooyman et al. [58] showed that tree height, wood density, leaf size and seed size was related to elevation in rainforest vegetation. Pollock et al. [78] showed that seed size in eucalypts in Victoria was related to topography and soil type, while Wills and Clarke [71] showed that trait types were associated with soil fertility.
Sampling in this study used PPBio for the purpose of long-term monitoring, therefore resampling of vegetation data will reoccur. This will give insight into vegetation change over time in relation to environmental variables, as demonstrated by [79] and [80]. Studying changes in tree cohort structure, species turnover and mortality may show that tree age structure, composition and abundance are related to dynamic variables such as weather patterns.
This study has shown that in a subtropical eucalypt forest patch, the tree assemblage patterns changed as a continuum along topographic and soil gradients at the mesoscale. Past research together with this study demonstrate that vegetation conservation at the assemblage level needs to consider the role of ecological processes at a variety of spatial scales. Mesoscale ecological studies also need to be replicated at a regional level to take into account the large distribution patterns of many vegetation assemblages and individual species. This would account for local and regional ecological effects and increase the power with which to elucidate the drivers of mesoscale vegetation patterns.
Acknowledgments
This study is part of a long-term project for the Planned Program for Biodiversity Studies (PPBio) and the TERN Australian SuperSite Network. Karawatha Forest Park is governed by the Brisbane City Council. We thank the PPBio Australia team who established the plots at Karawatha including Clay Simpkins, Daniel Stratford, Mikalah Malone, Subhashni Appanna, Kate Leopold-Wooldridge, Ryan Hughes, Saara Kampmann, Lucy Halliday and Joshua Morris, and the many volunteers from Griffith University. The assistance of Conservation Volunteers Australia in the field was very much appreciated, and Jutta Masterton and Andrew Bryant, and Michael Arthur provided statistical advice. Thanks also to the Queensland Herbarium staff and Kieran Richardt from Natura Consulting and Naomi Edwards for advice on species identification. We thank two anonymous reviewers and Paul Adam who provided insightful comments that improved the manuscript.
Author Contributions
Conceived and designed the experiments: J-MH JGC SAB. Performed the experiments: J-MH SAB. Analyzed the data: GWL SAB. Contributed reagents/materials/analysis tools: J-MH. Wrote the paper: GWL SAB J-MH JGC.
References
- 1. Austin MP (1985) Continuum concept, ordination methods and niche theory. Annual Review of Ecological Systems 16:39–61.
- 2. Collins SL, Glenn SM (1991) Importance of spatial and temporal dynamics in species regional abundance and distribution. Ecology 72:654–664.
- 3. Eiserhardt WL, Svenning J-C, Kissling WD, Balslev H (2011) Geographical ecology of the palms (Arecaceae): determinants of diversity and distributions across spatial scales. Annals of Botany 108:1391–1416.
- 4. Siefert A, Ravenscroft C, Althoff D, Avarez-Yépiz JC, Carter BE, et al. (2012) Scale dependence of vegetation-environment relationships: a meta-analysis of multivariate data. Journal of Vegetation Science 23:942–951.
- 5. Fensham RJ, Fairfax RJ, Dwyer JM (2012) Potential aboveground biomass in drought-prone forest used for rangeland pastoralism. Ecological Applications 22:894–908.
- 6. Batek MJ, Rebertus AJ, Schroeder WA, Haithcoat TL, Compas E, et al. (1999) Reconstruction of early nineteenth-century vegetation and fire regimes in the Missouri Ozarks. Journal of Biogeography 26:397–412.
- 7. Debski I, Burslem DFRP, Lamb B (2000) Ecological processes maintaining differential tree species distributions in an Australian subtropical rain forest: implications for models of species coexistence. Journal of Tropical Ecology 16:387–415.
- 8. Harms KE, Condit R, Hubbell SP, Foster RB (2001) Habitat associations of trees and shrubs in a 50-ha neotropical forest plot. Journal of Ecology 89:947–959.
- 9. Leathwick JR, Whitehead D (2001) Soil and atmospheric water deficits and the distribution of New Zealand's indigenous tree species. Functional Ecology 15:233–242.
- 10. Aguiar FC, Cerdeira JO, Martins MJ, Ferreira MT (2013) Riparian forests of Southwest Europe: are functional trait and species composition assemblages constrained by environment? Journal of Vegetation Science 24:628–638.
- 11. Nzegbule EC, Onyema MC, Ndelekwute SC (2013) Plant species richness and soil nutrients in a 35-year old cashew nut plantation in Isuochi, Southern Nigeria. Tropical Ecology 54:205–212.
- 12. Battaglia M, Williams KJ (1996) Mixed species stands of eucalypts as ecotones on a water supply gradient. Oecologia 108:518–528.
- 13. Partel M (2002) Local plant diversity patterns and evolutionary history at the regional scale. Ecology 83:2361–2366.
- 14. Sukri RS, Wahab RA, Salim KA, Burslem DFRP (2012) Habitat associations and community structure of Dipterocarps in response to environment and soil conditions in Brunei Darussalam, Northwest Borneo. Biotropica 44:595–605.
- 15. Olvera-Vargas M, Figueroa-Rangel BL, Vázquez-López JM (2010) Is there environmental differentiation in the Quercus-dominated forests of west-central Mexico? Plant Ecology 211:321–335.
- 16. Costa FR, Magnusson WE, Luizao RC (2005) Mesoscale distribution patterns of Amazonian understorey herbs in relation to topography, soil and watersheds. Journal of Ecology 93:863–878.
- 17. Clark DB, Palmer MW, Clark DA (1999) Edaphic factors and the landscape – scale distributions of tropical rainforest trees. Ecology 80:2662–2675.
- 18. Hutchinson TF, Boerner REJ, Iverson LR, Sutherland S, Sutherland EK (1999) Landscape patterns of understorey composition and richness across a moisture and nitrogen mineralisation gradient in Ohio (U.S.A) Quercus forests. Plant Ecology 144:177–189.
- 19. Ledo A, Burslem DFRP, Condés S, Montes F (2013) Micro-scale habitat associations of woody plants in a neotropical cloud forest. Journal of Vegetation Science 24:1086–1097.
- 20. Basnet K (1992) Effect of topography on the pattern of trees in Tabonuco (Decryodes excelsa) dominated rainforest of Puerto Rico. Biotropica 24:31–42.
- 21. Fensham RJ, Kirkpatrick JB (1992) Soil characteristics and tree species distributions in the Savanna of Melville Island, Northern Territory. Australian Journal of Botany 40:311–333.
- 22. Williams RJ, Duff GA, Bowman MJS, Cook GD (1996) Variation in the competition and structure of tropical savannas as a function of rainfall and soil texture along a large-scale climatic gradient in the Northern Territory, Australia. Journal of Biogeography 23:747–756.
- 23. O'Grady AP, Chen X, Eamus D, Hutley LB (2000) Composition, leaf area index and standing biomass of eucalypt open forests near Darwin in the Northern Territory, Australia. Australian Journal of Botany 48:629–638.
- 24. Pekin BK, Boer MM, Macfarlane C, Grierson PF (2009) Impacts of increased fire frequency and aridity on eucalypt forest structure, biomass and composition in southwest Australia. Forest Ecology and Management 258:2136–2142.
- 25. Pekin BK, Wittkuhn RS, Boer MM, Macfarlane C, Grierson PF (2012) Response of plant species and life form diversity to variable fire histories and biomass in the jarrah forest of south-west Australia. Austral Ecology 37:330–338.
- 26. Horton BM, Glen M, Davidson NJ, Ratkowsky D, Close DC, et al. (2013) Temperate eucalypt forest decline is linked to altered ectomycorrhizal communities mediated by soil chemistry. Forest Ecology and Management 302:329–337.
- 27. Catterall CP, Piper SD, Bunn SE, Arthur JM (2001) Flora and fauna assemblages vary with local topography in a subtropical eucalypt forest. Austral Ecology 26:56–69.
- 28. Hager T, Benson D (2010) The Eucalypts of the Greater Blue Mountains World Heritage Area: distribution, classification and habitats of the species of Eucalyptus, Angophora and Corymbia (family Myrtaceae) recorded in its eight conservation reserves. Cunninghamia 11:425–444.
- 29. Neave HM, Norton TW (1998) Biological inventory for conservation evaluation IV. Composition, distribution and spatial prediction of vegetation assemblages in southern Australia. Forest Ecology and Management 106:259–281.
- 30. Condit R, Sukumar R, Hubbell SP, Foster RB (1998) Predicting population trends from size distributions: a direct test in a tropical tree community. The American Naturalist 152:495–509.
- 31. Nathan R, Muller-Landau HC (2000) Spatial patterns of seed dispersal, their determinants and consequences for recruitment. Tree 15:278–285.
- 32.
McElhinny C (2002) Forest and woodland structure as an index of biodiversity: a review. Australian National University, Acton Act 0200: Department of Forestry, Australian National University.
- 33. Vlam M, Baker PJ, Bunyavejchewin S, Mohren GMJ, Zuidema PA (2014) Understanding recruitment failure in tropical tree species: Insights from a tree-ring study. Forest Ecology and Management 312:108–116.
- 34. Lai JS, Coomes DA, Du XJ, Hsieh CF, Sun IF, et al. (2013) A general combined model to describe tree-diameter distributions within subtropical and temperate forest communities. Oikos 122:1636–1642.
- 35.
Kordas G, Coutts RH, Catterall CP (1993) The Vegetation of Karawatha Forest and its Significance in the South East Queensland Landscape. Nathan: Griffith University.
- 36. Magnusson WE, Lima AP, Luizão R, Luizão F, Costa FRC, et al. (2005) RAPELD: A modification of the Gentry method for biodiversity surveys in long-term ecological research sites. Biota Neotropica 5:1–6.
- 37. Hero J-M, Castley JG, Malone M, Lawson B, Magnusson WE (2010) Long-term ecological research in Australia: innovative approaches for future benefits. Australian Zoologist 35:90–102.
- 38.
Magnusson WE, Braga-Neto R, Pezzini F, Landeiro VL, Baccaro F, et al.. (2013) Biodiversity and Integrated Environmental Monitoring. California, USA: Attema Design.
- 39. Hero J-M, Castley JG, Butler SA, Lollback GW (2013) Biomass estimation within an Australian eucalypt forest: meso-scale spatial arrangement and the influence of sampling intensity. Forest Ecology and Management 310:547–554.
- 40.
Queensland Herbarium (2001) Key to Eucalypts of greater Brisbane. Mt Coot-tha, Brisbane: Queensland Herbarium.
- 41.
Flora of Australia (1981) Flora of Australia. Volume 16 Elaeagnaceae, Proteaceae 1. Canberra: Australian Biological Resource Study.
- 42.
Australian Biological Resources Study (1995) Flora of Australia. Volume 16 Elaeagnaceae, Proteaceae 1. Canberra: Australian Biological Resources Study and CSIRO.
- 43.
Australian Biological Resources Study (2000) Flora of Australia. Volume 17a Proteaceae 2: Grevillea. Canberra: Australian Biological Resources Study and CSIRO.
- 44.
Autodesk (2006) Autocad <http://www.autodesk.com.au/>.
- 45.
Kalra YP, Maynard DG (1991) Methods manual for forest soil and plant analysis. Edmonton, Alberta: Forestry Canada, Northwest Region, Northern Forestry Centre.
- 46.
Rayment GE, Higginson FR (1992) The Australian handbook of soil and water chemical methods. Melbourne: Inkta Press.
- 47.
Legendre P, Legendre L (1998) Numerical ecology. Amsterdam: Elsevier.
- 48. Garden JG, McAlpine CA, Possingham HP, Jones DN (2007) Habitat structure is more important than vegetation composition for local-level management of native terrestrial reptile amd small mammal species living in urban remnants: A case study from Brisbane, Australia. Austral Ecology 32:601–719.
- 49.
R Core Team (2013) R: A language and environment for statistical computing <http://www.R-project.org/>.
- 50.
Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, et al.. (2013) vegan: Community Ecology Package <http://CRAN.R-project.org/package=vegan>.
- 51. Sklenar F, Ramsay PM (2001) Diversity of zonal paramo plant communities in Ecuador. Diversity and Distributions 7:113–124.
- 52. Austheim G (2002) Plant diversity patterns in semi-natural grasslands along an elevational gradient in southern Norway. Plant Ecology 161:193–205.
- 53. Bhattarai KR, Vetaas OR (2003) Variation in plant species richness of different life forms along a subtropical elevation gradient in the Himalayas, east Nepal. Global Ecology and Biogeography Letters 12:327–340.
- 54.
Nagy L, Grabherr G, Korner C, Thompson DBA (2003) Alpine biodiversity in space and time: a synthesis. In: Nagy L, Grabherr G, Korner C, Thompson DBA, editors. Alpine biodiversity in Europe. Berlin: Springer-Verlag.
- 55. Grytnes JA (2003) Species-richness patterns of vascular plants along seven altitudinal transects in Norway. Ecography 26:291–300.
- 56. Sanchez M, Pedroni F, Eisenlohr PV, Oliveira-Filho AT (2013) Changes in tree community composition and structure of Atlantic rain forest on a slope of the Serra do mar range, southeastern Brazil, from near sea level to 1000 m of altitude. Flora 208:184–196.
- 57. Zhang J-T, Xu B, Li M (2013) vegetation patterns and species diversity along elevational and disturbance gradients in the Baihua Mountain Reserve, Beijing, China. Mountain Research and Development 33:170–178.
- 58. Kooyman R, Cornwell W, Westoby M (2010) Plant functional traits in Australian subtropical rain forest: partitioning within-community from cross-landscape variation. Journal of Ecology 98:517–525.
- 59. Titus J (1990) Microtopography and woody plant generation in a hardwood floodplain swamp in Florida. Bulletin of the Torrey Botanical Club 117:429–437.
- 60. Olivaro-Filho AT, Vilela EA, Carvalho DA, Gavilanes ML (1994) Effects of soil and topography on the distribution of tree species in a tropical riverine forest in Southeastern Brazil. Journal of Tropical Ecology 10:483–508.
- 61. Austin MO (2002) Spatial prediction of species distribution: an interface between ecological theory and statistical modelling. Ecological Modelling 157:101–118.
- 62. Bledsoe BP, Shear TH (2000) Vegetation along hydrologic and edaphic gradients in North Carolina coastal plain creek bottom and implications for restoration. Wetlands 20:126–147.
- 63. Mark AF, Dickenson KJM, Hofstede RGM (2000) Alpine vegetation, plant distribution, life forms, and environments in a prehumid New Zealand region: oceanic and tropical high mountain affinities. Arctic, Antarctic, and Alpine Research 32:240–254.
- 64. Beerling DJ, Woodward FI, Lomas M, Jenkins AJ (1997) Testing the responses of a dynamic global vegetation model to environmental change: a comparison of observations and predictions. Global Ecology and Biogeography Letters 6.
- 65. Egan JL, Williams RJ (1996) Lifeform distributions of woodland species along a moisture availability gradient in Australia's monsoonal tropics. Australian Systematic Botany 9:205–217.
- 66. Bale CLW, Williams JB, Charley JL (1998) The impact of aspect on forest structure and floristics in some Eastern Australian sites. Forest Ecology and Management 110:363–377.
- 67. Liu X, Zhang W, Yang F, Zhou X, Liu Z, et al. (2012) Changes in vegetation-environment relationships over long-term natural restoration process in Middle Taihang Mountain of North China. Ecological Engineering 49:193–200.
- 68. Moeslund JE, Arge L, Bocher PK, Dalgaard T, Svenning J-C (2013) Topography as a driver of local terrestrial vascular plant diversity patterns. Nordic Journal of Botany 31:129–144.
- 69. Eisenlohr PV, Alves LF, Bernacci LC, Padgurschi CG, Torres RB, et al. (2013) Disturbances, elevation, topography and spatial proximity drive vegetation patterns along an altitudinal gradient of a top biodiversity hotspot. Biodiversity and Conservation 22:2767–2783.
- 70. Wardell-Johnson GW, Lawson BE, Coutts RH (2007) Are regional ecosystems compatible with floristic heterogeneity? A case study from Toohey Forest, south-east Queensland, Australia. Pacific Conservation Biology 13:47–59.
- 71. Wills KE, Clarke PJ (2008) Plant trait-environmental linkages among contrasting landscapes and climate regimes in temperate eucalypt woodlands. Australian Journal of Botany 56:422–432.
- 72. Pansonato MP, Costa FRC, De Castilho CV, Carvalho FA, Zuquim G (2013) Spatial scale or amplitude of predictors as determinants of the relative importance of environmental factors to plant community structure. Biotropica 45:199–307.
- 73. Austin MP, Smith TM (1989) A new model for the continuum theory. Vegetation 83:35–47.
- 74. Huston M, Smith T (1987) Plant succession: life history and competition. The American Naturalist 130:168–198.
- 75. Williams RJ, Cook GD, Gill AM, Moore PHR (1999) Fire regime, fire intensity and tree survival in a tropical savanna in northern Australia. Australian Journal of Ecology 24:50–59.
- 76. Westoby M, Falster DS, Moles AT, Vesk PA, Wright IJ (2002) Plant ecological strategies: some leading dimensions of variation between species. Annual Review of Ecology and Systematics 33:125–159.
- 77. Austin MP, Smith TM, Van Niel KP, Wellington AB (2009) Physiological responses and statistical models of the environmental niche: a comparative study of two co-occuring Eucalyptus species. Journal of Ecology 97:496–507.
- 78. Pollock L, Morris WK, Vesk PA (2012) The role of functional traits in species distributions revealed through a hierarchical model. Ecography 35:716–725.
- 79. de Castilho CV, Magnusson WE, De Araújo RNO, Luizão FJ (2010) Short-term temporal changes in tree live biomass in a central Amazonian forest, Brazil. Biotropica 42:95–103.
- 80. de Toledo JJ, Magnusson WE, Castilho CV, Nascimento HEM (2012) Tree mode of death in Central Amazonia: Effects of soil and topography on tree mortality associated with storm disturbances. Forest Ecology and Management 263:253–261.