Rising CO₂ :
(Publications 16,19,24,26,27,29,35,40,44,45,46,54,55,63,80,81,82,84,85,86,88,89,91,93)

 

Changes in atmospheric CO₂ concentration ([CO₂]) have direct and immediate effects on plant physiology, but as time progresses, a series of plant and ecosystem feedbacks act to modify this direct response. Our research has addressed both the short-term direct CO₂-response (24, 26, 40) and how it is modified by soil feedbacks (27, 35, 44, 45, 46, 86, 91). Different feedbacks have different time constants and hence become important on different timescales. For long-term predictions the primary scientific tools are ecosystem models based on an understanding of the processes involved, such as our model G'DAY (Generic Decomposition And Yield), which simulates the cycling of carbon (C), nitrogen (N) and water in plant and soil.

 

Research with G'DAY has investigated CO₂ effects on plants on different timescales and how responses at different times relate to each other. Our main focus has been on responses of trees, which are of particular interest because they grow to maturity on timescales comparable to the timescale of [CO₂] increase. Much of our modelling with G'DAY has been based on a mathematical technique called 'two-timing' or 'quasi-equilibrium' analysis. This analysis exploits the fact that ecosystems are made up of plant and soil pools with widely differing turnover times. For instance, leaves have faster turnover times than stemwood. Thus, on a specified timescale, the system may be separated into 'fast'- and 'slow'-turnover pools, where fast pools are in approximate (or quasi-) equilibrium and slow pools are effectively constant. At quasi-equilibrium the total flux of C into the fast pools is equal to the total efflux of C (and similarly for N). The C and N flux equilibria of the fast pools lead to an enormous simplification of G'DAY, which gives powerful insight into the model's behaviour. The analysis leads to explicit equations that explain in graphical fashion how the CO₂-response depends on all model parameters. The analysis has been applied to evaluate the response of NPP, the C sink and ecosystem C storage to rising [CO₂] for both N- and water-limited ecosystems, and has been used to identify key model parameters influencing responses on differing timescales (35, 45). Of particular importance are parameters describing the flexibility of plant and soil nitrogen to carbon (N:C) ratios; large responses occur if N:C ratios decline significantly at high [CO₂], with little or no response if N:C ratios are inflexible. If soil N:C ratios are assumed to be fixed, increased C flows to the soil must be accompanied by an increase in N immobilisation, limiting N availability for plants. However, if this ratio is variable, soil C storage may be increased without a concomitant reduction in N mineralisation. According to G'DAY, the CO₂?response changes over time because responses on longer timescales are dictated by the N:C ratios of less rapidly cycled organic matter.

 

Rising temperature:
(Publications 41, 42, 44, 46, 55, 64)

 

Research has addressed temperature effects on both plant physiology and soil processes.

 

Physiological acclimation to temperature:
Many published models predict that forests that are currently carbon sinks may become sources of C under increased temperature because photosynthesis (C uptake) is less temperature sensitive than respiration (C release). However, the temperature responses used in these models are obtained from short-term response curves; long-term, temperature-acclimated responses may be quite different. For example, recent experiments suggest that the whole-plant photosynthesis: autotrophic respiration ratio is in fact independent of temperature. We (Belinda & Ross, in collaboration with Roddy Dewar from INRA, Bordeaux) have investigated this issue by developing a mechanistic model of physiological acclimation to temperature, based on dynamics of substrate pools. A key prediction of the model is that the balance of respiration: photosynthesis (R:P), although strongly temperature-dependent when measured instantaneously, is independent of temperature on monthly time-scales (42).

 

Temperature effects on plant and soil:
Ross and Dave have modelled the effect of rising temperatures on productivity of both forests and grasslands (46, 55). Warming increases decomposition and soil N mineralisation, which stimulates net primary productivity (NPP) at all sites, except when inducing water limitation. At forest sites some of the enhanced N release is allocated to a woody biomass pool with a low N:C ratio so that warming significantly enhances carbon sequestration without increased N input at forest sites, but not at the grassland sites.

 

Acclimation of heterotrophic respiration to temperature:
Recent field measurements of soil respiration in forests and grasslands are leading scientists to question the traditional view that soil decomposition is a simple exponential function of soil temperature. Recent findings from soil warming experiments have shown that soil CO₂ efflux is stimulated by a step-increase in temperature, but declines under prolonged warming and eventually returns to rates similar to un-warmed soil. We (David & Ross in collaboration with Peter Eliasson & Professor Sune Linder from Swedish University of Agriculturla Sciences) have modelled this process for a soil-warming experiment in boreal forest (Flakaliden, Sweden) and showed that the observed acclimation of soil CO₂ flux can be largely explained by higher decomposition rates in warmed soil, leading to depletion of fast-turnover, labile soil organic matter pools (57).

 

Evidence of temperature effects from tree dendro-chronology:
Dave (in collaboration with Assoc. Prof. Brian Atwell at Macquarie University) has used a 200-year dendro-chronological record from old-growth Huon pine stands in western Tasmania to investigate how past changes in temperature have affected tree growth and nitrogen availability.

 

The above research on temperature effects is significant because of recent claims, based largely on the exponential temperature dependence of autotrophic and heterotrophic respiration, that the terrestrial biosphere, currently thought to constitute a gross carbon sink of approximately 2 billion tonnes per year, could become a carbon source within a few decades. Carbon models that ignore acclimation of respiration may therefore overestimate the future terrestrial carbon sink in response to global warming, while models that ignore soil feedbacks (such as several included in the IPCC Third Assessment Report 2001) may overestimate it.

 

Terrestrial carbon-sink saturation:
(Publications 46, 55)

 

One vitally important environmental issue is the future of the terrestrial carbon sink. Contrasting possibilities that have been suggested are: that the sink will decline over the next few decades due to soil C loss through temperature stimulation of soil respiration and saturation of the [CO₂] response, or that the sink will be sustained because growth increases due to CO₂ fertilisation and temperature stimulation of nutrient availability will override soil C losses. Dave and Ross have contributed to this debate by simulating the carbon sink of a Norway spruce stand at Flakaliden, northern Sweden, and two grassland ecosystems in USA, under various climate change scenarios (46, 55). Our work predicts that these systems will remain C sinks throughout the next century, but the size of the sink is sensitive to assumptions about rates of N deposition and variation in soil N:C ratios. Under rising [CO₂], simulations with flexible soil N:C ratio show evidence of C sink 'saturation', with net ecosystem production (NEP) increasing to a peak and then declining towards zero NEP.

 

Land-use change:
(Publications 47, 52, 53, 69, 96)

 

G'DAY has also been used to investigate effects of land-use change on soil carbon (C) storage, which globally constitutes a large component of biospheric carbon storage. An important question is whether soil C is gained or lost following land-use change. We reviewed the literature to assess changes in soil C upon conversion of forests to agricultural land; conversion of forest to cultivated land led to an average soil-C loss of approximately 30%, whereas under conversion of forest to pasture there was no significant change (47). In both transitions the change in soil C was positively correlated with soil-N change. Two past Honours students (Joanne Halliday & Liz Heagney) have also conducted field and modelling studies on changes in soil C following afforestation of pastureland, in collaboration with New Zealand Landcare and CSIRO Forestry (52, 53).

 

 

Our current work on the above topics includes: