An ecophysiological basis for the assembly of Australian rainforest tree communities

doi:10.21203/rs.3.rs-3833899/v1

Download PDF

Article

An ecophysiological basis for the assembly of Australian rainforest tree communities

https://doi.org/10.21203/rs.3.rs-3833899/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Understanding how past climate has filtered different tree strategies into communities is crucial for predicting how future climates will impact species and communities, yet few studies have used physiologically interpretable traits to explain the assembly of entire tree communities across large, continuous climatic gradients. To address this gap, we systematically surveyed rainforest tree communities across the Australian subtropics (spanning 600 to 2,500 mm rainfall yr^− 1) and measured functional traits on 285 (91%) of the recorded tree species, including detailed measurements of xylem anatomy to describe species’ hydraulic strategies. The direction and shape of species’ occurrence trends across the regional moisture gradient were strongly related to their hydraulic strategies. Evergreen species with efficient hydraulics were more prevalent in mesic locations, while those with safer hydraulics favoured drier climates. Despite having extremely efficient hydraulics, deciduous species declined along the moisture gradient. At the dry end of the gradient, lower soil fertility increased the prevalence of very safe evergreen strategies and decreased the prevalence of deciduous species, relative to high-fertility sites. Overall, we reveal how climate, soil and biogeography have jointly filtered tree strategies into communities across the Australian subtropics, providing a general foundation for prediction under ongoing climate change.

Biological sciences/Ecology/Community ecology

Biological sciences/Ecology/Ecophysiology

Biological sciences/Ecology/Forest ecology

functional traits

habitat filtering

hydraulic strategies

leaf economic spectrum

angiosperms

dispersal

Rainforests are estimated to support over 70% of the world’s terrestrial biodiversity¹ and contribute strongly to global carbon and water cycles^2,3. Ongoing anthropogenic climate change poses a considerable threat to rainforest communities, already driving widespread tree mortality, species redistribution and reductions in tropical rainforest canopy cover^4,5. Prediction of longer-term climate change impacts demands a deeper and more general understanding of how past climate has filtered different tree strategies into contemporary assemblages.

Hydraulic strategies are likely to be important drivers of tree species distributions along gradients of moisture availability^6,7. In drier tropical forests, evergreen species exhibit strong embolism resistance and low hydraulic efficiency (K_s) allowing them to tolerate extended dry periods^8,9, while deciduous species maintain high K_s during favourable periods and avoid dry periods by shedding leaves and storing water in sapwood^10,11. In mesic rainforests, evergreen species typically have less safe hydraulic strategies than those in dry forests¹², but light-driven trade-offs create considerable strategy variation¹³. Shade-tolerant understory species exhibit low K_s and invest in wood and leaf tissue to maximise survival^14,15. Conversely, pioneer species usurp transient light resources in gaps via high K_s and photosynthetic capacity¹⁴ and low stem and leaf construction costs¹⁶, but have shorter lifespans. Long-lived canopy species exhibit intermediate strategies, requiring high K_s to meet transpirational demands¹⁶ and overcome the hydraulic resistance associated with long xylem path lengths¹⁷, while also investing in structural support¹⁸.

Physiological studies have revealed mechanistic traits directly related to environmental tolerances^7,9. However, due to the expense and difficulty of measuring direct physiological traits¹⁹, such studies tend to be conducted on small, potentially biased subsets of species. Comprehensively characterising the strategies of entire assemblages in many locations necessitates a shift towards easier-to-measure traits, ideally without sacrificing a lot of physiological resolution. For this purpose, soft traits with well demonstrated links to underlying physiological processes^7,20, and integrative traits related to how species deal with major limiting resources along gradients of interest^21,22, are required. Despite their potential, such integrative traits have not yet been used to explain the assembly of rainforest tree assemblages across large-scale climate gradients.

In this study we systematically surveyed assemblages of rainforest trees across the Australian subtropics (spanning 600 to 2500 mm rainfall yr^− 1; Fig. 1), and comprehensively measured traits describing the hydraulic strategies of over 90% of the recorded angiosperm species (Table S1). We fit trait-by-environment interactions in a generalised linear mixed effects model to quantify how our chosen traits modulate species’ occurrence probabilities along climate and soil gradients^23,24. We hypothesised that functional traits describing hydraulic strategies strongly modulate tree species’ occurrence probabilities along the region’s pronounced moisture availability gradient. For a given climate, we also hypothesised that sites with more fertile and finely textured soils support more hydraulically efficient species than sites with less fertile, coarse-textured soils^25,26. Finally, we expected to find signals of Pleistocene climatic oscillations modulated by species’ hydraulic and dispersal traits. Specifically, we hypothesised that sites far from known climate refugia²⁷ contain species with safer hydraulic strategies and smaller diaspores than sites closer to refugia²⁸.

Principal component analysis (PCA) of seven climate variables identified a clear and dominant axis of moisture availability on PC1 (Fig. 2a & Extended Data Table 1), while PCA of seven continuous trait variables identified a hydraulic safety-efficiency spectrum on PC1, a leaf economic spectrum on PC2 (Fig. 2b) and a H_max axis on PC3 (a residual height axis after removing small H_max contributions to PC1 and PC2; Extended Data Fig. 1 & Extended Data Table 2).

The GLMM had a marginal R² of 0.33. The model allowed each species to have its own quadratic occurrence relationship with Climate PC1, which comprehensively captured species’ realised niches along the dominant climate gradient (Fig. 3), as demonstrated by the high conditional R² of 0.88.

As hypothesised, the interaction between Climate PC1 and Trait PC1 was by far the strongest, and in general, interactions involving either Trait PC1 or Climate PC1 were stronger than other interactions (Extended Data Fig. 2 & Extended Data Table 3). Species that increased along Climate PC1 had more efficient hydraulics (high Trait PC1 values; Fig. 4a), were taller (Fig. 4b), were evergreen (Fig. 4c) and had larger diaspores (Fig. 4d) than species that declined along Climate PC1. The Climate PC1 quadratic term also interacted strongly with Trait PC1, modulating the shape of species’ occurrence trends along the moisture gradient (Extended Data Figs. 2 & 3a). After accounting for Climate PC1, species with efficient hydraulics were more likely to occur at higher relative elevations than at lower relative elevations (Extended Data Fig. 3b), consistent with occult precipitation inputs in elevated positions.

Plant-available P was involved in a series of interactions that described two counteracting effects. First, a relatively weak, negative interaction between trait PC1 and plant-available P (Extended Data Fig. 2) indicated that hydraulically efficient species tended to decline with increasing P (seemingly counter to our second hypothesis), while safe strategies had no consistent trends with P (Extended Data Fig. 3c). Second, a stronger, positive interaction between Climate PC1 and plant-available P (Extended Data Fig. 2) captured effects of plot-level stem densities on average occurrence probabilities. In dry locations, overall occurrence probabilities (and stem densities) were highest in low-fertility sites and declined sharply with increasing plant-available P (Extended Data Fig. 4). In wet locations an opposing but weaker trend was evident. The combined result of these interactions (Fig. 5) was much higher occurrence probabilities for very safe evergreen strategies in dry, low-fertility sites compared with dry, high-fertility sites, which is generally consistent with our second hypothesis. At the mesic end of the climate gradient, effects of P on the occurrence probabilities of different hydraulic strategies were relatively minor (Fig. 5).

In partial support of our third hypothesis, species with safe hydraulics were more likely to occur further from climate refugia (Extended Data Fig. 3d), but diaspore size interacted only weakly with distance to refugia (refer to full model summary in Table S4). Deciduousness also interacted with plant-available P and distance to refugia (Extended Data Fig. 2), such that deciduous species were more likely to occur in sites with higher plant-available P (Extended Data Fig. 3e) and further from climate refugia (Extended Data Fig. 3f). There were other significant, but weaker, interactions between traits and environmental variables in the GLMM (Extended Data Fig. 2).

This study combined extensive survey data and comprehensive trait measurements on almost 300 tree species to reveal the ecophysiological basis for how climate, soil and biogeography have structured rainforest tree assemblages across 9° of latitude in the Australian subtropics. The filtering effect of climate was overarching; the interaction between Trait PC1 and Climate PC1 was more than twice as strong as any other trait-by-environment interaction that we examined (Extended Data Fig. 2). In fact, based on the strength of this interaction and others, the model reproduced the community weighted means for all traits very reliably (Extended Data Fig. 5).

Consistent with the widely reported safety-efficiency trade-off, species with highly efficient water-use strategies were more prevalent in mesic locations (Fig. 4a), permitting greater transpiration and growth rates required to reach higher-radiation positions in the forest canopy, and achieve their large reproductive size^16,29. Furthermore, the inclusion of a deciduous indicator variable captured drought-avoiding strategies as viable alternatives to safe evergreen strategies in drier climates (Figs. 4a & b), consistent with findings in tropical forests^30,31. While such traits relating to tree hydraulics are known to influence species’ distributions along climatic gradients^32,33, we provide an unparalleled, community-scale demonstration of how species’ hydraulic traits modulate their occurrences along a large and continuous climate gradient.

The hydraulic axis also interacted significantly with the quadratic Climate PC1 term (Extended Data Fig. 2), whereby species with safer strategies tended to have linear relationships (quadratic coefficients closer to zero) and efficient species had more hump-shaped trends (more negative quadratic coefficients) along the moisture gradient (Extended Data Fig. 3a). The more linear relationships for very safe species arose for two reasons. First, the distributions of very safe species extend into drier climates³⁴, so our data did not capture how their occurrence probabilities decline beyond our study region. Second, many very safe species occurred in all (or most) very dry sites, allowing their probabilities of occurrence to decline linearly (on the modelled scale) from high values in very dry locations, to lower values as moisture availability increased. The hump-shaped trends for hydraulically efficient species likely emerged for the opposite reason. Very mesic rainforests support more species, including more endemic and range-restricted tree species²⁷ thus reducing the average likelihood of recording a given efficient species in multiple mesic sites (Extended Data Fig. 6).

Occurrence probabilities of efficient hydraulic strategies also increased with relative elevation (Extended Data Fig. 3b), especially at the wet end of Climate PC1, where rainforests at higher relative elevations were more likely to contain species with highly efficient hydraulics compared with those at low relative elevations (not shown). This can likely be attributed to the region’s topographic complexity which positions a number of our sites near the orographic cloud base³⁵. Consistent with the Massenerhebung effect, small, isolated peaks (with greater relative elevations) support a lowering of the cloud base, associated with higher relative atmospheric humidity at a given altitude^36,37. Fog or cloud may further improve plant water status by reducing solar radiation and thus water loss via transpiration, as well as by facilitating direct foliar uptake of water³⁸. This effect can decouple a site’s microclimate from the regional climate and buffer against climatic extremes³⁹, in turn supporting tree species with greater hydraulic demands.

With increasing moisture availability (along Climate PC1) we also captured the emergence of tall canopy species (Fig. 4d) with relatively small, reinforced lumens (as indicated by Trait PC3). While taller trees have access to high-radiation habitats in mesic forests, they must also tolerate increasing vessel path length resistance^17,40, as well as greater vapour pressure deficits in exposed forest canopies⁴¹. Consequently, in these taller canopy trees, pronounced vessel tapering may serve to reduce cavitation risk in longer, inherently more vulnerable vessels – preventing large negative tensions in the most distal portions of the canopy⁴². Indeed, this is consistent with hydraulic theory⁴³ and has been observed in tropical tree species⁴⁴. Ultimately, we demonstrate here that tall canopy trees, while employing efficient water-use strategies, can coordinate their traits to circumvent the biophysical challenges of larger stature⁴².

These taller species at the positive extreme of Trait PC3 were also more likely to occur in locations with lower minimum temperatures (Extended Data Fig. 2), which were typically high-altitude sites far from the coast (not shown). Such cooler, elevated sites are less likely than lower-elevation sites to experience the combined effects of drought and extreme heat^45,46, which may enable the persistence of larger trees by minimizing the risk of drought-induced hydraulic failure^47,48. Our results are mirrored in mesic, tropical forests; Binks et al.⁴⁹ identified that, under nonlimiting soil water supply, rainforest tree communities growing under higher temperatures tended to be shorter in stature.

Our model also revealed strong trait-modulated responses of species to soil fertility and texture. Compared to their evergreen counterparts, deciduous species were more likely to occur in higher-P sites (Extended Data Fig. 3e), which aligns with findings in tropical dry forests^50,51. Drought-deciduous species typically combine high hydraulic efficiency with acquisitive leaf economics to enable rapid re-leafing and to maximise growth during favourable climatic periods²⁹. As such, these species require higher soil fertility to facilitate the frequent turnover of high-SLA leaves, with nutrient loss via leaf shedding not adequately compensated in low-fertility soils⁵², even if some nutrient resorption occurs prior to shedding⁵³. Consequently, evergreen species that combine safe hydraulics with conservative leaf economics become dominant in less fertile dry forests^30,54,55, as we found in this study (Fig. 5). Safe, conservative strategies also confer greater resistance to physical damage, preventing high repair costs in infertile sites^56,57.

Soil texture is expected to operate as an additional edaphic filter by moderating the total amount of plant-available soil water in these rainforest systems²⁵. The lower water-holding capacity of coarse-textured soils⁵⁸ led species to adopt more conservative leaf economics (higher Trait PC2 values) (Extended Data Fig. 2), likely to help maintain leaf turgor at lower leaf water potentials^9,20. In contrast, species with ‘fast’ leaves (lower Trait PC2 values) were associated with clay soils that improve soil water access⁵², consistent with higher rates of resource acquisition in more productive habitats⁵⁹. These results indicate that soil fertility and texture modulate the adaptive value of traits in a given climate, and thus underscore the importance of accounting for soil, climate and their interactive effects as habitat filters during community assembly⁵⁴.

In addition to strong climatic and edaphic filtering, we found strong signatures of dispersal and biogeographic processes that have also structured contemporary rainforest assemblages. At the mesic extreme of Climate PC1, higher-altitude sites have been identified previously as regional refugia – areas that have maintained stable rainforest habitat through historical climate oscillations, most recently during the Pleistocene^27,60. We accounted for these refugia by simply measuring the distance of each site to the closest refugium (as identified by Weber et al.²⁷) and then adjusting for correlation with climate variables to separate historical from contemporary processes. While it has long been recognized that these refugia support a greater richness of endemic species⁶¹, our data shows that these stable refugia have specifically permitted the persistence and accumulation of evergreen species with efficient hydraulic strategies (Extended Data Fig. 3b). Within these habitats, light competition is expected to have driven strong life history trade-offs among species, reinforcing selection for highly efficient water-use strategies that permit taller canopy positions^62,63. Moreover, we revealed that deciduous species most commonly occur further from these refugial locations (Extended Data Fig. 3f), likely because deciduousness is a less competitive strategy in environments where water is reliably available⁶⁴.

The distance to refugia-by-diaspore diameter interaction was only close to significant in our full model (Table S4), but two other strong and significant interactions indicate that dispersal has also played an important role in structuring tree communities during Pleistocene climatic oscillations^28,65. The Climate PC1-by-diaspore diameter (Fig. 4c) and relative elevation-by-diaspore diameter interactions (Extended Data Fig. 3i) both indicated that larger-fruited species increase in occurrence probability as the climate becomes wetter, and as relative elevation increases. In a supplementary analysis (Supplementary information), we examined species’ dispersal modes using a mixed-effects multinomial model and found that occurrence probabilities of species with large endozoochorous diaspores (> 10 mm) increased sharply along Climate PC1 (Extended Data Fig. 7), while those with small diaspores declined just as sharply. This is consistent with well-dispersed, smaller-fruited species recolonising less stable habitats during wetter interglacial periods²⁸, and also aligns with the probable extinction of large frugivores from the subtropics that has since inhibited dispersal of larger-fruited species out of refugia^66,67.

The strength of trait-by-climate interactions in our model highlights its considerable potential for predicting future distributions of tree strategies, or even entire assemblages, under climate change. We refrained from making such projections using available ensemble models because of high uncertainty in rainfall predictions for the Australian subtropics⁶⁸. In ongoing research, we will explore applications of our data to storylines that more transparently represent such uncertainty⁶⁹. Nevertheless, this study has revealed how climate, soil and biogeography have jointly acted on tree hydraulic, leaf economic and dispersal strategies to create the distinct assemblages present across the Australian subtropics. Our strong results emphasise the importance of choosing functional traits that are related to the major gradients of resource availability⁷⁰, and with well-understood links to relevant physiological processes^21,22, to increase the physiological resolution of community assembly research.

Study system

This study was undertaken in subtropical rainforest and dry forest between Gladstone in Central Queensland and Ballina in New South Wales, Australia, spanning nine degrees of latitude (Fig. 1). Subtropical rainforests and dry forests are ideal for studying climate-driven forest assembly, as they span large moisture gradients, but unlike tropical rainforests or sclerophyll forests, they are not frequently impacted by large-scale disturbances such as cyclones or fire (but see^71,72). Consequently, the signal of climate on assembly is expected to be strong.

At the time of European colonisation, rainforest was patchily distributed throughout the subtropics of Australia (Fig. 1). These forests form distinct assemblages across the region’s climate gradients and have been classified into subformations based on structural and physiognomic features, including; canopy height and strata, leaf size, and the proportion of evergreen, semi-evergreen, deciduous or Araucarian emergent trees⁷³. These subformations are broadly correlated with gradients of temperature, moisture and soil properties (e.g. fertility and texture)⁷⁴.

Survey design and methodology

We surveyed 45 rainforest communities (i.e. sites), distributed continuously across the region’s major climate gradients (Fig. 1). Sampled communities were located on soils derived from basalt or granite⁷⁵, the most widespread substrates that support rainforest in the region. Within each site, four 25 m x 25 m plots were established across 1–2 ha of homogeneous forest, and the stem diameter (DBH) and species identity of all self-supporting woody individuals ≥ 3 cm were recorded. Plots were spaced at least 50 m apart. We avoided large canopy gaps and areas with signs of historical logging to best represent climax assemblages across the region.

Environmental variables

Climate variables

We extracted climate data from Harwood et al.⁷⁶ and ANUCLIMATE 2.0⁷⁷ for each site location (mean plot latitude and longitude) using the extract() function from the terra package⁷⁸. We chose mean annual vapour pressure deficit (hPa) and mean annual frost days from ANUCLIMATE 2.0⁷⁷, and precipitation seasonality (equinox seasonality), mean T_max (^oC), mean T_min (^oC), mean annual moisture index (mean annual precipitation/ mean annual potential evaporation) and mean annual precipitation (mm) from Harwood et al.⁷⁶. Mean annual frost days was square-root-transformed, and mean annual precipitation and mean annual moisture index were ln-transformed to approximate normal distributions. These seven variables were subjected to principal component analysis (PCA) to produce two orthogonal climate axes for subsequent statistical analyses. Climate PC1 (60.8%) was loaded by mean annual precipitation (+), mean annual moisture index (+), vapour pressure deficit (-), precipitation seasonality (-) and mean T_max (-) and represented a gradient of moisture availability and timing across the study region. Climate PC2 (34.4%) was loaded by mean annual frost days (-) and T_min (+) and captured a gradient of minimum temperature (Fig. 2a, Extended Data Table 1).

To capture the potential drying and warming effect of each plot’s topographical position, we calculated the amount of direct solar radiation (Watt/m²) received by each plot using slope and aspect data. Using available 5-m contour information⁷⁹ we calculated the slope and aspect of each plot in QGIS⁸⁰. From this, we used the DirectRadiation() function from the solrad package⁸¹ to determine the mean annual solar radiation at 3 pm as an indicator of exposure to drying conditions.

Soil variables

Soil properties are known to influence community composition via edaphic compensation, with finer-textured and fertile soils supporting species with more acquisitive traits²⁵. Topsoil samples (10–20 cm depth) were collected from each plot corner, mixed thoroughly, and air-dried in the laboratory at The University of Queensland. Samples were sent for analysis to CSBP (Soil and Plant Analysis Laboratory, Bibra Lake, Western Australia) to measure plant-available macronutrients. We selected plant-available (Colwell) phosphorous (P, mg/kg) to represent soil fertility, as P is a major limiting resource⁸² and has been linked to turnover in rainforest tree strategies^30,54. P was moderately positively correlated with Climate PC1. To prevent multicollinearity in subsequent statistical models, we regressed ln-transformed P on Climate PC1 and Climate PC2 and extracted the residuals to remove correlation with climate. While P was collected and measured at the plot level, we used site-level means in subsequent statistical analyses to represent average site soil fertility. We also extracted clay content for each site from the Soil and Landscape Grid of Australia⁸³, with site clay content averaged across 0–5, 5–15 and 15–30 cm depths.

Context variables

The context of a site in the surrounding landscape or region is known to influence the diversity and composition of plant communities⁸⁴. Given the patchy pre-European distribution of subtropical rainforests, we accounted for potential patch size effects^85,86 by estimating the area of pre-1750 rainforest⁸⁷ surrounding each site. Specifically, we used the extract() function from the terra package⁷⁸ to sum up the total number of rainforest pixels (1-ha cell size) within a 5-km radius of each site, where “rainforest” was defined as NVIS (National Vegetation Information System) major vegetation subgroups 2 (tropical or subtropical rainforest), 6 (warm temperate rainforest) and 62 (dry rainforest or vine thicket)⁸⁷.

Climate oscillations during the Pleistocene also influenced the diversity, composition and phylogenetic structure of rainforest tree communities in the Australian subtropics^27,28,60. Pleistocene oscillations resulted in the contraction of rainforest to stable refugia during glacial periods, followed by the re-expansion of rainforest during wetter interglacial periods. Regions that maintained rainforest habitat since the last glacial maximum (LGM) (i.e. refugia) tend to have higher diversity than more contemporary rainforest expansions, including a greater richness of range-restricted endemic species²⁷. To represent the distance of each site to historical rainforest refugia²⁷, we manually measured the distance from each plot to the centre of the closest refugium in QGIS⁸⁰ and averaged these distances for each site. Refugia included Lamington NP, Blackall Range and Bulburin NP (Fig. 1).

We also described the topographic position of each site relative to rainforest in the surrounding area to capture multiple context-dependent processes. These processes include cloud stripping that contributes large additional water inputs at high relative elevations⁸⁸, as well as potential species pool and dispersal effects. For example, a high-elevation site surrounded mainly by lower elevation rainforest may include more species from lower elevations than expected due to mass effects⁸⁹. We masked a digital elevation model (DEM)⁹⁰ by the NVIS pre-1750 rainforest extent, extracted all elevation values within a 5-km radius of each site and calculated each site’s quantile within the distribution of surrounding elevation values. Values of 1 corresponded to the highest position within the surrounding rainforest and values of 0 corresponded to the lowest position. A 5-km radius was chosen to calculate the surrounding rainforest area and relative elevation variables because it is a feasible distance over which species could have dispersed in recent ecological time^65,91, while being large enough to capture major topographic variation in the surrounding landscape (e.g. from mountain tops to adjacent valleys). Smaller and larger radii were examined but performed similarly or worse as explanatory variables in statistical models.

Like the soil P variable, two context variables (distance to the closest refugium and surrounding rainforest area) were moderately correlated with Climate PC1. To prevent multicollinearity, we removed climate correlation using simple regressions as aforementioned. Both variables were first sqrt-transformed to approximate normal distributions.

Trait sampling and measurement

We measured morphological and anatomical traits that are well-understood proxies for direct, physiological processes of tree hydraulic and light utilisation strategies across hundreds of species (Table S1). Branches and sun-exposed leaf samples were collected from wild and cultivated adult specimens throughout the south-east Queensland region. So long as sampled trees were healthy and sun-exposed branches were collected, trait values did not differ systematically between wild and cultivated individuals.

Lamina area, SLA and LDMC

For each species, measurements were taken on sets of leaves from four different individuals. To prevent desiccation, cut branches were stored in wetted plastic bags during transport to the laboratory. To rehydrate leaves, branches were recut in deionised water and covered with a black plastic bag until the following day. Four undamaged, fully expanded leaves were selected from each branch following rehydration and weighed with an analytical balance (GR-200, A&D Company Ltd, Tokyo, Japan) to attain fresh mass, then scanned through a leaf area meter (LI-3100C Area Meter, LI-COR Biosciences, Lincoln, USA) to measure lamina area. For compound leaves, half of the leaflets were selected, with measurements performed on each individual leaflet after it had been separated from the leaf rachis. After drying leaves at 70°C for three days, specific leaf area (SLA; lamina area / dry mass; mm²/mg) and leaf dry matter content (LDMC; dry mass / fresh mass; proportion) were calculated for each individual leaf or leaflet.

Petiole vessel traits

We used microscopy on petioles to measure xylem lumen breadth (b) and vessel wall thickness (t). Species-level averages for b and (t/b) were calculated to describe hydraulic efficiency and vessel reinforcement, respectively²². Refer to Supplementary Information for details.

Wood density

Wood density (dry mass / fresh volume; g cm^− 3) was measured on branch samples (5–7 cm long, 2–7 cm diameter) from three individuals of each species. After removing the bark, the water displacement method was used to measure fresh volume. Dry mass was then attained after samples were dried for three days at 105°C.

H_max and deciduousness

Deciduousness and species’ maximum heights (H_max) were obtained from Harden et al.⁹², and revised where necessary based on observations during surveys. We acknowledge that there are several forms of deciduousness⁹³, but given the small number of deciduous species in our dataset (18/285 species, ~ 6%), and the information about species’ hydraulic and leaf economic strategies captured by other traits, we used a simple binary classification in this study.

Diaspore diameter

We included diaspore diameter to capture species’ differing abilities to recolonise rainforest patches during wetter interglacial periods. While this variable does not capture the mode of dispersal, endozoochory is by far the most common mode, and the most variable in diaspore size (Fig. S1). As such, diaspore diameter separated well-dispersed, small-fruited species from larger-fruited, endozoochorous species that are likely more restricted. Refer to Supplementary Information for more details.

Creating orthogonal strategy axes

Individual trait values were averaged by species, which was supported by variance components analyses confirming that > 70% of the variation in all continuous traits was among species (Fig. S2). SLA, diaspore diameter, b and t/b were ln-transformed, and H_max and lamina area were square-root-transformed to approximate normal distributions. Three gymnosperms (two Araucaria and one Podocarpus species) that lack xylem vessels were excluded from trait analyses, as were two hemi-parasitic species (Exocarpus latifolius and Santalum lanceolatum). To generate orthogonal strategy axes for statistical analyses, the PCA() function from the FactoMineR package⁹⁴ was applied to standardized trait values (mean of 0, SD = 1) of all recorded species with complete trait data (285 species). Deciduousness and diaspore diameter were not included in the trait PCA, but they were included as separate trait variables in subsequent statistical models.

Trait PC1 (48%) was most strongly loaded by b (+), lamina area (+), wood density (-) and t/b (-), with all of these traits loading greater than |0.43|. Thus, Trait PC1 captured a hydraulic safety-efficiency spectrum, whereby negative Trait PC1 scores indicate safe hydraulic strategies (dense wood, small leaves and narrow, reinforced vessel lumens) and high scores indicate more efficient hydraulic strategies (Fig. 2b). Trait PC2 (22.1%) was dominated by SLA (-) and LDMC (+), representing the leaf economics spectrum (LES) from acquisitive to conservative strategies (Fig. 2b & Extended Data Table 2). H_max loaded only moderately on the first two principal components; positively with more efficient strategies on Trait PC1 and positively with LDMC on Trait PC2. Trait PC3 (10.9%) was strongly loaded by H_max (+), capturing the residual effect of height after accounting for its minor contributions to Trait PC1 and Trait PC2 (Extended Data Table 2 & Extended Data Fig. 1). Trait PC3 was also moderately loaded by t/b (+), b (-), and wood density (-), indicating that taller species on this axis had relatively small, well-reinforced vessels for their stature.

Statistical modelling

All analyses were performed in R version 4.2.3⁹⁵. The response variable was binary indicating the presence (1) or absence (0) of all 285 species in each of the 180 plots (51,300 observations in total). We modelled the occurrence of species $j$ in plot $i$ within site $k$, $\text{P}\text{r}\left({Y}_{ijk}=1\right)$ in a Bernoulli generalised linear mixed-effects model (GLMM) with logit link:

where ${p}_{ijk}$ is the probability of occurrence; ${\alpha }_{ik}$ and ${\beta }_{0j}$ are plot-within-site and species random intercepts, respectively. The parameters ${\beta }_{jm}$ capture the response of species $j$ to the $m$th environmental explanatory variable ${X}_{mik}$, while ${\theta }_{n}$ captures the main effects of the $n$th species’ traits ${T}_{jn}$. We included eight environmental explanatory variables describing climate, soil and context (Table S2) plus one quadratic term of Climate PC1. The environmental explanatory variables are denoted by subscripts $i$ and $k$ to imply that they were measured either at the plot or site level. Next, we included interactions between species’ traits ${T}_{jn}$ and environmental variables ${X}_{mik}$ to understand how traits modulate species’ environmental responses through the fourth-corner coefficients ${\kappa }_{mn}$. Our goal here can be better understood by rewriting the trait-by-environment terms above as:

$$\sum _{m=1}^{9}{\beta }_{jm}{X}_{mik}+\sum _{m=1}^{9}\sum _{n=1}^{5}{\kappa }_{mn}{X}_{mik}{T}_{jn}=\sum _{m=1}^{9}\left({\beta }_{jm}+\sum _{n=1}^{5}{\kappa }_{mn}{T}_{jn}\right){X}_{mik}\hspace{0.17em},$$

which shows how traits ${T}_{jn}$ may modulate the species’ environmental responses ${\beta }_{jm}$. Lastly, we included interactions between Climate PC1 and three selected environmental variables, with ${Z}_{lik}$ denoting the three-column design matrix of these interactions for brevity and ${\gamma }_{l}$ denoting their corresponding interaction coefficients.

Continuous explanatory variables were standardized (mean = 0, SD = 1) to facilitate model convergence, except for trait and climate PC variables that were already centred on zero. Refer to Table S4 for the initial “full” model that was fitted. We conducted one round of model simplification by dropping all non-significant two-way interactions.

Models were fit using the glmmTMB() function from the glmmTMB package⁹⁶. Model diagnostics were checked using the simulateResiduals() function from the DHARMa package⁹⁷. The marginal and conditional R² value was estimated using the r2() function from the performance package⁹⁸. To visualise trait-by-environment interactions, we first calculated species’ partial slopes for a given environmental variable and plotted these (on the Y-axis) against the relevant trait variable (on the X-axis). We then plotted the corresponding trait-by-environment relationship using the main effect for the environmental variable as the intercept and the relevant interaction term as the slope⁹⁹.

In a pilot analysis (Supplementary Information), we fitted a basic phylogenetic GLMM to examine any evolutionary signal in species’ responses to Climate PC1 and its quadratic term, i.e., two of the ${\beta }_{jm}$. There was, however, only weak phylogenetic signal in these species’ slopes (Table S3), so phylogeny was excluded from subsequent analyses.

Acknowledgements

We acknowledge the Turrabal, Yagera, Yugambah, Wakka Wakka, and Gabi Gabi People as the Traditional Owners on whose land much of this research took place. We also acknowledge the many Indigenous groups across Queensland and New South Wales where field surveys were undertaken and pay tribute to their historic and continued sovereignty and connection to Country. This research was funded by the Hermon Slade Foundation (HSF20050). We would like to thank Joseph Douglas, Jacob White, Tom Hanley and Ashleigh Ford for their assistance with rainforest field surveys, and Alison Brown and Don Butler for providing tree survey and trait data for the Bunya Mountains. Thanks to Bill McDonald, Spencer Shaw, Rhonda Green, Cath Moran and Don Butler for imparting their expert knowledge of Australia’s subtropical rainforests, and Queensland Parks and Wildlife Service, Sunshine Coast Regional Council, NSW National Parks and Wildlife Service and private landholders that facilitated access to the study sites. We are also grateful to Daniel Falster, Isaac Towers, Mark Westoby, Tim Brodribb, David Keith, Andy Pitman, Angela Moles, Stephen Bonser and Kasia Ziemińska for their useful advice and suggestions.

Author contributions

JRS collected all field and trait data with assistance from ECVW and JMD. JRS led measurement of leaf and wood traits while ECVW led microscopy for vessel traits. JRS and JMD analysed data with guidance from HRL. JRS drafted the manuscript with guidance from JMD and HRL. All authors contributed to editing and improving the manuscript.

Reaka-Kudla, M. L. in Biodiversity II: Understanding and Protecting Our Biological Resources Vol. 2 (eds M.L. Reaka-Kudla, D.E. Wilson, & E.O. Wilson) Ch. 7, 549 (Joseph Henry Press, 1996).
Malhi, Y. & Grace, J. Tropical forests and atmospheric carbon dioxide. TREE 15, 332-337 (2000).
Reichstein, M. et al. Climate extremes and the carbon cycle. Nature 500, 287-295 (2013).
Lyra, A. et al. Projections of climate change impacts on central America tropical rainforest. Clim. Change 141, 93-105 (2016).
Allen, C. D. et al. A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For. Ecol. Manage. 259, 660-684 (2010).
Engelbrecht, B. M. et al. Drought sensitivity shapes species distribution patterns in tropical forests. Nature 447, 80-82 (2007).
Laughlin, D. C. et al. Climatic limits of temperate rainforest tree species are explained by xylem embolism resistance among angiosperms but not among conifers. New Phytol 226, 727-740 (2020).
Bartlett, M. K., Detto, M. & Pacala, S. W. Predicting shifts in the functional composition of tropical forests under increased drought and CO2 from trade-offs among plant hydraulic traits. Ecol. Lett. 22, 67-77 (2019).
Markesteijn, L., Poorter, L., Paz, H., Sack, L. & Bongers, F. Ecological differentiation in xylem cavitation resistance is associated with stem and leaf structural traits. Plant Cell Environ. 34, 137-148 (2011).
Larcher, W. Physiological Plant Ecology: Ecophysiology and Stress Physiology of Functional Groups. Vol. 3 (Springer, 2003).
Meinzer, F. C., Johnson, D. M., Lachenbruch, B., McCulloh, K. A. & Woodruff, D. R. Xylem hydraulic safety margins in woody plants: Coordination of stomatal control of xylem tension with hydraulic capacitance. Funct. Ecol. 23, 922-930 (2009).
Tavares, J. V. et al. Basin-wide variation in tree hydraulic safety margins predicts the carbon balance of Amazon forests. Nature 617, 111-117 (2023).
Wright, I. J. et al. Functional traits and the growth-mortality trade-off in tropical trees. Ecology 91, 3664-3674 (2010).
Poorter, L. & Bongers, F. Leaf traits are good predictors of plant performance across 53 rain forest species. Ecology 87, 1733-1743 (2006).
King, D. A. Influence of light level on the growth and morphology of saplings in a Panamanian forest. Am. J. Bot. 81, 948-957 (1994).
Poorter, L. et al. The importance of wood traits and hydraulic conductance for the performance and life history strategies of 42 rainforest tree species. New Phytol 185, 481-492 (2010).
Ryan, M. G. & Yoder, B. J. Hydraulic limits to tree height and tree growth. BioScience 47, 235-242 (1997).
Niklas, K. J. Plant Biomechanics: An Engineering Approach to Plant Form and Function. (The University of Chicago Press, 1992).
Funk, J. L. et al. Revisiting the Holy Grail: Using plant functional traits to understand ecological processes. Biol. Rev. 92, 1156-1173 (2017).
Mendez-Alonzo, R., Paz, H., Zuluaga, R. C., Rosell, J. A. & Olson, M. E. Coordinated evolution of leaf and stem economics in tropical dry forest trees. Ecology 93 (2012).
Blackman, C. J., Brodribb, T. J. & Jordan, G. J. Leaf hydraulic vulnerability is related to conduit dimensions and drought resistance across a diverse range of woody angiosperms. New Phytol 188, 1113-1123 (2010).
Blackman, C. J. et al. The links between leaf hydraulic vulnerability to drought and key aspects of leaf venation and xylem anatomy among 26 Australian woody angiosperms from contrasting climates. Ann. Bot. 122, 59-67 (2018).
Laughlin, D. C. et al. Root traits explain plant species distributions along climatic gradients yet challenge the nature of ecological trade-offs. Nat. Ecol. Evol 5, 1123-1134 (2021).
Jamil, T., Ozinga, W. A., Kleyer, M. & ter Braak, C. J. F. Selecting traits that explain species–environment relationships: A generalized linear mixed model approach. J. Veg. Sci. 24, 988-1000 (2012).
Curran, T. J., Clarke, P. J. & Warwick, N. W. M. Water relations of woody plants on contrasting soils during drought: Does edaphic compensation account for dry rainforest distribution? Aust. J. Bot. 57, 629-639 (2010).
Webb, L. J. & Tracey, J. G. in Ecological Biogeography of Australia, Monographie Biologicae Vol. 41 (ed A. Keast) 605-694 (W Junk, 1981).
Weber, L. C. et al. Patterns of rain forest plant endemism in subtropical Australia relate to stable mesic refugia and species dispersal limitations. J. Biogeogr. 41, 222-238 (2014).
Kooyman, R., Rossetto, M., Allen, C. & Cornwell, W. K. Australian tropical and subtropical rain forest community assembly: Phylogeny, functional biogeography, and environmental gradients. Biotropica 44, 668-679 (2012).
Markesteijn, L., Poorter, L., Bongers, F., Paz, H. & Sack, L. Hydraulics and life history of tropical dry forest tree species: Coordination of species' drought and shade tolerance. New Phytol 191, 480-495 (2011).
Oliveira, R. S. et al. Linking plant hydraulics and the fast–slow continuum to understand resilience to drought in tropical ecosystems. New Phytol 230, 904-923 (2021).
Santiago, L. S. et al. Coordination and trade-offs among hydraulic safety, efficiency and drought avoidance traits in Amazonian rainforest canopy tree species. New Phytol 218, 1015-1024 (2018).
Fortunel, C., Paine, C. E. T., Fine, P. V. A., Kraft, N. J. B. & Baraloto, C. Environmental factors predict community functional composition in Amazonian forests. J. Ecol. 102, 145-155 (2014).
Blanchard, G. et al. Regional rainfall and local topography jointly drive tree community assembly in lowland tropical forests of New Caledonia. J. Veg. Sci. 30, 845-856 (2019).
Forster, P. I., Bostock, P. D., Bird, L. H. & Bean, A. R. Vineforest plant atlas for South-East Queensland with assessment of conservation status. (Queensland Herbarium, 1991).
Narsey, S. et al. Impact of climate change on cloud forests in the Gondwana Rainforests of Australia World Heritage Area. (NESP Earth Systems and Climate Change Hub, Australia, 2020).
Foster, P. The potential negative impacts of global climate change on tropical montane cloud forests. Earth Sci. Rev. 55, 73-106 (2001).
Bruijnzeel, A., Waterloo, M. J., Proctor, J., Kuiters, A. T. & Kotterink, B. Hydrological observations in montane rain forests on Gunung Silam, Sabah, Malaysia, with special reference to the `Massenerhebung' effect. J. Ecol. 81, 145-167 (1993).
Yates, D. & Hutley, L. Foliar uptake of water by wet leaves of Sloanea woollsii, an Australian subtropical rainforest tree. Aust. J. Bot. 43, 157-167 (1995).
Dobrowski, S. Z. A climatic basis for microrefugia: The influence of terrain on climate. Glob. Chang. Biol. 17, 1022-1035 (2011).
Koch, G. W., Sillett, S. C., Jennings, G. M. & Davis, S. D. The limits to tree height. Nature 428, 851-854 (2004).
Vinod, N. et al. Thermal sensitivity across forest vertical profiles: Patterns, mechanisms, and ecological implications. New Phytol 237, 22-47 (2023).
Couvreur, V. et al. Water transport through tall trees: A vertically explicit, analytical model of xylem hydraulic conductance in stems. Plant Cell Environ. 41, 1821-1839 (2018).
Olson, M. E. & Rosell, J. A. Vessel diameter–stem diameter scaling across woody angiosperms and the ecological causes of xylem vessel diameter variation. New Phytol 197, 1204-1213 (2013).
Schuldt, B., Leuschner, C., Brock, N. & Horna, V. Changes in wood density, wood anatomy and hydraulic properties of the xylem along the root-to-shoot flow path in tropical rainforest trees. Tree Physiol 33, 161-174 (2013).
Bauman, D. et al. Tropical tree mortality has increased with rising atmospheric water stress. Nature 608, 528-533 (2022).
Fontes, C. G. et al. Dry and hot: The hydraulic consequences of a climate change–type drought for Amazonian trees. Phil. Trans. R. Soc. B 373 (2018).
McDowell, N. G. et al. Pervasive shifts in forest dynamics in a changing world. Science 368 (2020).
Liu, L. et al. Tropical tall forests are more sensitive and vulnerable to drought than short forests. Glob. Chang. Biol. 28, 1583-1595 (2022).
Binks, O. J. et al. Vapour pressure deficit modulates hydraulic function and structure of tropical rainforests under nonlimiting soil water supply. New Phytol 240, 1405-1420 (2023).
Russo, S. E. & Kitajima, K. in Tropical Tree Physiology: Adaptations and Responses in a Changing Environment (eds Guillermo Goldstein & Louis S. Santiago) 357-383 (Springer International Publishing, 2016).
Condit, R., Engelbrecht, B. M., Pino, D., Perez, R. & Turner, B. L. Species distributions in response to individual soil nutrients and seasonal drought across a community of tropical trees. PNAS 110, 5064-5068 (2013).
Ouedraogo, D. Y. et al. The determinants of tropical forest deciduousness: Disentangling the effects of rainfall and geology in central Africa. J. Ecol. 104, 924-935 (2016).
Rentería, L. Y., Jaramillo, V. J., Martínez-Yrízar, A. & Pérez-Jiménez, A. Nitrogen and phosphorus resorption in trees of a Mexican tropical dry forest. Trees 19, 431-441 (2005).
Simpson, A. H., Richardson, S. J. & Laughlin, D. C. Soil–climate interactions explain variation in foliar, stem, root and reproductive traits across temperate forests. Glob. Ecol. Biogeogr. 25, 964-978 (2016).
McDonald, P. G., Fonesca, C. R., Overton, J. & Westoby, M. Leaf-size divergence along rainfall and soil-nutrient gradients: Is the method of size reduction common among clades? Funct. Ecol. 17, 50-57 (2003).
Poorter, L. et al. Are functional traits good predictors of demographic rates? Evidence from five neotropical forests. Ecology 89, 1908-1920 (2008).
Curran, T. J., Gersbach, L. N., Edwards, W. & Krockenberger, A. K. Wood density predicts plant damage and vegetative recovery rates caused by cyclone disturbance in tropical rainforest tree species of North Queensland, Australia. Austral Ecol. 33, 442-450 (2008).
Kramer, P. & Boyer, J. Water Relations of Plants and Soils. (1995).
Reich, P. B. The world-wide ‘fast–slow’ plant economics spectrum: A traits manifesto. J. Ecol. 102 (2014).
Kooyman, R., Rossetto, M., Cornwell, W. K. & Westoby, M. Phylogenetic tests of community assembly across regional to continental scales in tropical and subtropical rain forests. Glob. Ecol. Biogeogr. 20, 707-716 (2011).
Heads, M. Globally basal centres of endemism: The Tasman-Coral Sea region (south-west Pacific), Latin America and Madagascar/South Africa. Biol. J. Linn. Soc. Lond. 96, 222-245 (2008).
Zhang, J., Nielsen, S. E., Mao, L., Chen, S. & Svenning, J. Regional and historical factors supplement currentclimate in shaping global forest canopy height. J. Ecol. 104, 271-610 (2016).
Liu, H. et al. Hydraulic traits are coordinated with maximum plant height at the global scale. Sci. Adv. 5, 1-14 (2019).
Vico, G., Dralle, D., Feng, X., Thompson, S. & Manzoni, S. How competitive is drought deciduousness in tropical forests? A combined eco-hydrological and eco-evolutionary approach. Environ. Res. Lett. 12 (2017).
Butler, D. W., Green, R. J., Lamb, D., McDonald, J. F. & Forster, P. I. Biogeography of seed-dispersal syndromes, life-forms and seed sizes among woody rain-forest plants in Australia's subtropics. J. Biogeogr. 34, 1736-1750 (2007).
Rossetto, M. & Kooyman, R. The tension between dispersal and persistence regulates the current distribution of rare palaeo-endemic rain forest flora: A case study. J. Ecol. 93, 906-917 (2005).
Rossetto, M., Kooyman, R., Sherwin, W. & Jones, R. Dispersal limitations, rather than bottlenecks or habitat specificity, can restrict the distribution of rare and endemic rainforest trees. Am. J. Bot. 95, 321-329 (2008).
Dowdy, A. et al. East Coast Cluster Report. (CSIRO and Bureau of Meteorology, Australia, 2015).
Shepherd, T. G. et al. Storylines: An alternative approach to representing uncertainty in physical aspects of climate change. Clim. Change 151, 555-571 (2018).
Sanaphre-Villanueva, L. et al. Above- and below-ground trait coordination in tree seedlings depend on the most limiting resource: A test comparing a wet and a dry tropical forest in Mexico. PeerJ 10 (2022).
Kooyman, R., Watson, J. & Wilf, P. Protect Australia's Gondwana Rainforests. Science 367, 1083 (2020).
Olsen, M. & Lamb, D. Recovery of subtropical rainforest following storm damage. Proc. Ecol. Soc. Aus. 15, 197-301 (1988).
Webb, L. J. A physiognomic classification of Australian rain forests. J. Ecol. 47, 551-570 (1959).
Webb, L. J. Environmental relationships of the structural types of Australian rain forest vegetation. Ecology 49, 296-311 (1968).
Wilson, P. R. & Taylor, P. M. Land Zones of Queensland. 79 (Queensland Herbarium, Queensland Department of Science, Information Technology, Innovation and the Arts, Brisbane, 2012).
Harwood, T. et al. Climatology for continental Australia 1976-2005: Summary variables with elevation and radiative adjustment. (2016).
Hutchinson, M., XU, T., Kesteven, J., Marang, I. & Evans, B. ANUClimate v.2.0., (NCI Australia, Canberra, ACT, Australia, 2022).
Hijmans, R. J. raster: Geographic Data Analysis and Modeling. https://CRAN.R-project.org/package=raster (2021).
Queensland Government. Elevation contours - 10 metre interval - by area of interest. (Department of Resources, 1970).
QGIS Geographic Information System (QGIS Association, 2023).
Seyednasrollah, B., Kumar, M. & Link, T. E. On the role of vegetation density on net snow cover radiation at the forest floor. Geophys. Res. Atmos. 118, 8359–8374 (2013).
Hou, E. et al. Global meta-analysis shows pervasive phosphorus limitation of aboveground plant production in natural terrestrial ecosystems. Nat. Commun. 11, 637 (2020).
Viscarra Rossel, R. et al. Clay (3" resolution) (ed CSIRO) (2014).
Hille Ris Lambers, J., Adler, P. B., Harpole, W. S., Levine, J. M. & Mayfield, M. M. Rethinking community assembly through the lens of coexistence theory. Annu. Rev. Ecol. Evol. Syst. 43, 227-248 (2012).
Munguia-Rosas, M. A. & Montiel, S. Patch size and isolation predict plant species density in a naturally fragmented forest. PLoS One 9, e111742 (2014).
MacArthur, R. H. & Wilson, E. O. The equilibrium theory of island biogeography. (Princeton University Press, 1967).
Australian Government. in NVIS Australia - Pre-1750 Major Vegetation Subgroups NVIS Version 6.0 (Albers 100m analysis product) (ed Energy Australian Government Department of Climate Change, the Environment and Water) (2020).
Hutley, L. B., Doley, D., Yates, D. J. & Boonsaner, A. Water balance of an Australian subtropical rainforest at altitude: The ecological and physiological significance of intercepted cloud and fog. Aust. J. Bot. 45, 311-329 (1997).
Leibold, M. A. et al. The metacommunity concept: A framework for multi-scale community ecology. Ecol. Lett. 7, 601-613 (2004).
Gallant, J., Ticket, P. K., Wilson, N., Dowling, T. & Read, A. 1 second SRTM Level 2 Derived Smoothed Digital Elevation Model (DEM-S). (2010).
Mokany, K., Prasad, S. & Westcott, D. A. Loss of frugivore seed dispersal services under climate change. Nat. Commun. 5, 3971 (2014).
Harden, G., McDonald, W. J. F. & Williams, J. Rainforest Trees and Shrubs - Second Edition. (Gwen Harden Publishing, 2018).
Eamus, D. Ecophysiological traits of deciduous and evergreen woody species in the seasonally dry tropics. TREE 14, 11-16 (1999).
Le, S., Josse, J. & Husson, F. FactoMineR: An R package for multivariate analysis. J. Stat. Softw. 25, 1-18 (2008).
R: A language and environment for statistical computing (R Foundation for Statistical Computing, Vienna, Austria, 2021).
Brooks, M. et al. glmmTMB balances speed and flexibility among packages for zero-inflated generalized linear mixed modeling. R. J. 9 (2017).
DHARMa: Residual Diagnostics for Hierarchical (Multi-Level / Mixed) Regression Models (2022).
Ludecke, D. et al. performance: An R package for assessment, comparison and testing of statistical models. JOSS 6 (2021).
Gelman, A. & Hill, J. Data analysis using regression and multi-level/hierarchical models. (Cambridge University Press, 2007).

There is NO Competing Interest.

RadfordSmithetalsuppinformation040124.docx
Supplementary Information
ExtendedData.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

An ecophysiological basis for the assembly of Australian rainforest tree communities

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Study system

Survey design and methodology

Environmental variables

Climate variables

Soil variables

Context variables

Trait sampling and measurement

Lamina area, SLA and LDMC

Petiole vessel traits

Wood density

H_max and deciduousness

Diaspore diameter

Creating orthogonal strategy axes

Statistical modelling

Declarations

Acknowledgements

Author contributions

References

Additional Declarations

Supplementary Files

Status:

Version 1

An ecophysiological basis for the assembly of Australian rainforest tree communities

Status:

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Study system

Survey design and methodology

Environmental variables

Climate variables

Soil variables

Context variables

Trait sampling and measurement

Lamina area, SLA and LDMC

Petiole vessel traits

Wood density

Hmax and deciduousness

Diaspore diameter

Creating orthogonal strategy axes

Statistical modelling

Declarations

Acknowledgements

Author contributions

References

Additional Declarations

Supplementary Files

Status:

Version 1

H_max and deciduousness