The eco-physiology of two contrasting arid-zone woodlands in Australia

Publication Type:
Issue Date:
Full metadata record
Semi-arid and arid ecosystems occupy 45 % of the Earth’s land surface and approximately 40 % of the global population live in arid and semi-arid regions. Ecosystem productivity in these regions is constrained by water availability, which is in general erratic, spatially variable and confined to short periods during the wet season. Globally, semi-arid and arid ecosystems dominate the inter-annual variability of the global land carbon (C) sink. In particular, Australian semi-arid and arid regions were estimated to account for 60 % of the 2011 global land carbon sink anomaly (GLSA). Despite the importance of these Australian arid and semi-arid environments, mechanisms that explain variability in rates of C uptake (at regional- and global-scales) and trends are poorly understood, and these ecosystems remain little studied. Australia is an extensive and flat continent, of which 70 % is arid or semi-arid land. Two biomes dominate the central semi-arid region: (1) Mulga woodland, dominated by species of the genus Acacia (shallow rooted N₂-fixing trees, from the Mulga complex of species); and (2) open Corymbia savanna where the dominant cover is spinifex (a C4 grass) with widely spaced tall evergreen Corymbia trees. These two ecosystems are found within the Ti-Tree Basin, Northern Territory Australia, where two eddy covariance systems has been in operation for the past 4 and 7 years within a Mulga woodland and Corymbia savanna respectively. The main objective of this research was to investigate ecosystem functioning of the two semi-arid woodlands, in order to improve our understanding of the interaction of terrestrial semi-arid ecosystems with the atmosphere, through measurements of C and water fluxes at ecosystem- and leaf-scales. The overall hypothesis was that differences in ecosystem vegetation composition and structure would be significant factors explaining differences in C and water fluxes across two disparate ecosystems. To accomplish this general objective, C and water fluxes were evaluated at different temporal-scales (i.e., diurnal, seasonal, annual and inter-annual) and different spatial-scales (from plot- to leaf-scales) within a Mulga woodland and a Corymbia savanna. I employed three different approaches to evaluate C and water fluxes: i) eddy covariance data (at plot-scale); ii) a range of in situ eco-physiological investigations (at leaf-scale); and iii) glasshouse experimentation (at leaf- and whole-plant-scale). In 2011 during the GLSA, the Mulga woodland captured 131 g C m⁻² y⁻¹ and total annual precipitation was 565 mm. The most recent hydrological year studied (August to July 2016-2017) had the largest annual rainfall recorded during my monitoring of ecosystem fluxes (713 mm) and net ecosystem production (NEP) was 217 gC m⁻² y⁻¹. In contrast to the Mulga woodland, the open Corymbia-savanna was a C source across most years (2012 to 2016), with NEP ranging between -14 and -190 gC m⁻² y⁻¹, but was a C sink during 2016-2017 with NEP of 115 gC m⁻² y⁻¹. As a result of continuous monitoring of C fluxes, precipitation thresholds at which the two semi-arid woodlands switched from C source to C sink were identified for the first time. The pivot-point for the Mulga woodland was 262 mm y⁻¹ and 506 mm y⁻¹ for the Corymbia savanna. The two semi-arid woodlands experience the same climatic conditions; hence, I observed that different climatic drivers (i.e., temperature, vapour pressure, soil water content) exerted similar influences over C and water fluxes across seasons. Intra-annual variability in C and water fluxes was mostly attributed to differences in SWC across seasons. However, different eco-physiological behaviours of co-occurring species within the Mulga woodland and Corymbia savanna contributed to explain differences in C and water fluxes between them. Dominant species at the Mulga woodland are highly adapted to periods of low water availability. Thus Mulga species were very water-use-efficient (WUE: trade-off between C gain per water loss) compared to species at the Corymbia savanna. This was demonstrated when evaluating responses of the g₁ parameter (as a proxy of intrinsic water-use-efficiency; WUEᵢ) to water availability. The Mulga woodland (at ecosystem- and leaf-scale) demonstrated large capacity for water consumption in wet periods and the plasticity to become highly WUE when experiencing water scarcity. In contrast, dominant species at the Corymbia savanna had larger water use (i.e., large stomatal conductance, at ecosystem-scale large rates of evapotranspiration) thus, smaller WUE compared to species at the Mulga woodland. The C4 spinifex grass can be highly productive at the Corymbia savanna during wet periods and, it is likely that the biomass produced during 2010-2011 was the fuel for photo-degradation of leaf litter, particularly of spinifex leaves in subsequent years, and this can explain the large negative NEP observed at the Corymbia savanna for much of the study period. One of the novel aspects of the present research was to evaluate how soil water content drives WUE using the theory of the optimal stomata behaviour through the g₁ parameter (a normalized metric of intrinsic water-use efficiency: ᵢWUE) and compered g₁ values estimated from three different methods (leaf gas exchange, carbon isotopes and eddy covariance fluxes). Here my result showed discrepancies among methods and seasons. Thus, Seasonal and annual variation in g₁ derived from the three methods in this present study highlights the concern that the generic use of constant values of g₁ to describe stomatal functioning is not reliable when parameterizing global climate models (Medlyn et al., 2017; Wolz et al., 2017). My research at a leaf-scale highlighted the importance of species-specific attributes in driving C fluxes in semi-arid Australia. Integrating plant eco-physiological responses of dominant species is an essential step for improving our understanding of C flux rates within a Mulga woodland and Corymbia savanna. Comparing semi-arid ecosystems contributes to our understanding of ecosystem functioning and mechanisms underlying variability in rates of C and water flux within different ecosystems, which bring us closer to understanding global variability in C cycling in terrestrial ecosystems. The research at a leaf-scale highlighted the importance of species-specific attributes of co-occurring species driving C fluxes in central Australia. Understanding functional processes (i.e., C assimilation, stomatal behaviours plant water status) and the vegetation responses across different plant species and ecosystems is crucial for improving our ability to predict global change. There is now an opportunity to evaluate the inclusion of in situ observations and account for variations of C and water fluxes when applying earth systems models and terrestrial vegetation C models at regional- and global-scales.
Please use this identifier to cite or link to this item: