Skip to main content

Does rapid utilisation of elevated nutrient availability allow eucalypts to dominate in the tropical savannas of Australia?

Cite this dataset

Paramjyothi, Harinandanan et al. (2020). Does rapid utilisation of elevated nutrient availability allow eucalypts to dominate in the tropical savannas of Australia? [Dataset]. Dryad.


Northern Australia's savannas are amongst the most fire-prone biomes on Earth, and are dominated by eucalypts (Eucalyptus and Corymbia spp.). It is not clear what processes allows this group to dominate under such extreme fire frequencies and if a superior ability to compete for nutrients and water might play a role. There is evidence that eucalypts are adapted to frequent fires; juvenile eucalypts escape the fire trap by growing rapidly in height between fires. However, non-eucalypts are less able to escape the fire trap and tend to have stand structures strongly skewed towards suppressed juveniles. The mechanisms that drive these contrasting fire responses are not well understood. Here we describe the results of a controlled glasshouse seedling experiment that evaluated the relative importance of nutrient and water availability in determining height growth and biomass growth of two eucalypt and one non-eucalypt tree species, common in northern Australian savannas. We demonstrate that growth of eucalypt seedlings is particularly responsive to nutrient addition. Eucalypt seedlings are able to rapidly utilise soil nutrients and accumulate biomass at a much greater rate than non-eucalypt seedlings. We suggest that a seasonal spike in nutrient availability creates a nutrient rich microsite that allows eucalypt seedlings to rapidly gain height and biomass, increasing their likelihood of establishing successfully and reaching a fire-resistant size. Our results extend our understanding of how eucalypts dominate northern Australian savannas under extremely high fire frequencies.


Study species

Three tree species were chosen for this study, two eucalypts (Eucalyptus tetrodonta and Eucalyptus miniata, family Myrtaceae) and one non-eucalypt (Erythrophleum chlorostachys, family Fabaceae). These are three of the most abundant tree species throughout northern Australia’s mesic savannas (mean annual rainfall >1000 mm) (Figure 2). Across an extensive array of vegetation monitoring sites in the mesic savannas of Kakadu, Nitmiluk and Litchfield National Parks, near Darwin, Eucalyptus tetrodonta contributes, on average, 16% of total tree basal area, Eucalyptus miniata 16% and Erythrophleum chlorostachys 13% (Russell-Smith et al. 2010).

Eucalyptus miniata and Eucalyptus tetrodonta are considered to be the most dominant trees in the northern Australian savanna woodlands (Russell-Smith et al 2010). Both species usually grow to 15-20 m in height but can also grow up to 30 m in favourable conditions (Setterfield 1997). Both species regenerate mainly through vegetative reproduction (lignotubers and, for Eucalyptus tetrodonta, occasionally from root sprouts) and from seed (Lacey & Whelan 1976; Setterfield 1997). Erythrophleum chlorostachys is the dominant non-Eucalypt species in northern Australian savanna tree canopies, but also dominants in the sub canopy layer (Setterfield 1996, 1997; Russell-Smith et al 2010). Erythrophleum chlorostachys grows to a maximum height of 18 m under optimum conditions and is a nitrogen fixing legume species (Fensham & Bowman 1995). The study species naturally grow in the savanna woodlands of northern Australia where the soils are typical of sandy or sandy loam, extensively weathered and laterised, climate is wet-dry tropical, with approximately 95% of the annual rainfall (~1700 mm year-1) occuring during the wet season (November to April) in Darwin (the study location) (Chen et al 2005).

Seedling establishment

The experiment was undertaken in an outdoor shade house near Darwin, Australia in 2010. Seeds from Eucalyptus tetrodonta, Eucalyptus miniata and Erythrophleum chlorostachys were germinated in seedling trays (each cell 3 × 3 cm, 7 cm depth) filled with a soil mixture consisting of equal volume of washed river sand and coconut peat. Erythrophleum chlorostachys seeds were lightly scarified with coarse sandpaper prior to planting. Seeds were placed in shallow depressions in the soil surface and partially covered with soil. Seedlings were grown in seedling trays in a shade house under ambient temperatures and sufficient water to prevent the growth medium from drying out for 11–14 weeks. Seedlings were then transplanted into tall pots (10 cm diameter × 40 cm depth) containing natural topsoil, sourced from savanna woodland near Darwin that naturally supports the study species, mixed with approximately 20% fine sand, by volume.

Experimental design

Four treatments were applied to the seedlings. These treatments were designed to simulate environmental stressors faced by tree species in northern Australian savannas: moisture and nutrient limitation. A fully-crossed factorial design was applied with seven replicates of each species in each treatment level. There was (1) an ambient water and no additional nutrients (Ambient), (2) Ambient water and additional nutrients (N+), (3) Additional water and no additional nutrients (W+) and (4) Additional water and additional nutrients (W+N+) treatments. For the N+ treatment, 5 g of all-purpose fertilizer (Osmocote all-purpose fertilizer (Scotts Australia, Everris, The Netherlands); N (20.9): P (0.5): K (3.8)) was added to each pot. As this is a slow release fertilizer (12 months of longevity), we applied a smaller amount of fertilizer per pot (1.5g/L) than the recommended usage (5g/L). Additional nutrients were added only once, at the beginning of the experiment. For the W treatment, watering occurred twice each day for 1.5 min (3min/day), and for the W+ treatment, four times per day (6 min/day). Watering occurred at a rate of 0.22 mL cm-2 min-1. Compared to the average annual rainfall of the study site (~1700mm year-1), this watering treatments were high (~2000mm year-1 for W treatment and ~4000mm year-1 for W+ treatment). Extrapolation of the watering treatments to annual rainfall was done for four months (December- March) as 90% of the rainfall occuring in this period of time in Darwin region (Bureau of Meteorology, Australia). Stored water in the seedling pots was minimal due to drainage and the extra water treatment compensated this loss. We used a mist spray to water the plants so that the plants were not drenched or saturated during watering.

Growth metrics and statistical analysis

Height of each seedling was measured every 14 days from the beginning of the experiment. Height was measured from the cotyledon scar to the growing tip. The experiment was terminated after six months, all seedlings were harvested and biomass of each seedling was separated into leaves, stems, roots and lignotubers. The mass of each component was measured after oven-drying the samples at 80°C for 48 hours.

We examined four seedling response variables, each measured at the end of the experiment: height; root:shoot ratio; total biomass; and belowground biomass. All response variables were log-transformed prior to analysis to ensure normality. Generalized linear modelling was used to investigate the effects of the experimental treatments on each seedling response variable (R package glm2; Marschner 2014). The three species were analysed separately.

The best model of each seedling response variable was selected using Akaike’s Information Criterion (AIC) (Table S1 & S2). We began the model selection with a saturated model, containing two binary predictor variables (W+ treatment, N+ treatment, and their interaction) and then iteratively removed predictor variables, searching for the model with the lowest AIC. If an individual predictor did not add sufficient explanatory power relative to its contribution to model complexity, as defined by AIC, it was dropped from the model. All analyses were performed in RStudio version 3.4.1 (RStudio 2015) and the R package “ggplot2” (Wickham 2016) used for diagrams.


The Hermon Slade Foundation, Award: HSF15/07