SUSTAINABLE PRODUCTIVITY OF MANAGED FORESTS
(Publications 37, 38, 39, 48, 49, 69, 70, 71, 90)

 

In Australian and global forestry, there is increasing emphasis on planting trees on short rotations for pulpwood. Several studies have shown however that fast-growing eucalypt plantations can rapidly exhaust soil water and/or nutrient reserves, which raises concerns about the environmental impact and viability of short-rotation forestry in infertile or irregular rainfall areas. Our research in this area has aimed to develop methodologies for evaluating sustainability under different management practices, and for identifying best management practices. This research pioneered the use of ecosystem models as tools for evaluating ecological sustainability.

 

Our initial models of sustainability (developed with Dr Roddy Dewar, now with INRA, Bordeaux) were based on a theoretical analysis of long-term changes in forest growth as a result of ecosystem N loss associated with forest management practices. We considered a forest stand undergoing a repeated management cycle with forest harvests followed by slash fires and re-planting. Over successive rotations, soil N supply may decrease due to N removal by harvesting, fire, leaching and gaseous emissions. In the long-term the system approaches a steady state at which total N losses equal N inputs over a rotation. We defined sustainable yield as the mean annual stemwood volume increment (MAI) achieved at that steady state, and developed a graphical method for evaluating the steady state, based on the balance between N inputs and outputs. Using this method sustainable yield can be evaluated without the need to run simulations of forest growth over multiple rotations.

 

The second phase of this research (Ross & Dave in collaboration with Marc Corbeels, then with CSIRO, Perth, and now with CIRAD, Brazil), involved modelling the sustainability of short-rotation Eucalyptus globulus plantations in south-western Australia, where vast areas of eucalypt plantations have been established on naturally infertile ex-pasture soils that have been improved through fertilisation and planting of legume-based pastures for grazing of sheep and cattle. High soil nutrient supply rates mean that tree growth rates are initially high. However, plantations are managed on short rotations (e.g. 10 years), and recent research indicates that this change in land use may, in the longer term, lead to a decline in availability of some plant nutrients, especially nitrogen and phosphorus. Studies have measured lower levels of soil net N mineralisation and soil inorganic N under mature E. globulus stands than under pasture (though both may be elevated in the early stages of stand development when N mineralisation is surplus to plant demand). These decreases raise the question of whether the high productivity achieved by first rotation E. globulus stands planted on improved pasture land will be maintained in subsequent rotations. We compared our simulations of G'DAY with the earlier graphical method for quantifying sustainable forest yield, and found that simulations followed a similar pattern with MAI and rotation-averaged N supply both declining over successive rotations. We compared simulations for sites of contrasting fertility and inter-rotation management practices, (including harvest-residue retention, fertilisation and fire practices), and found that the rate of MAI decline was highest and the sustainable (steady- state) MAI was lowest at infertile, low rainfall sites with removal of harvest residues.

 

We have also used G'DAY to investigate why soil N availability often declines following conversion of pasture to forest plantations. Possible explanations for the decline include: reduced N input under forest, e.g. through cessation of N fixation, or cessation of grazing; enhanced N leaching in the early years of stand development; increased litter input to soil and/or reduced forest-litter quality, leading to immobilisation of soil N; increased N removals in forest harvests and fires; a transfer of soil N reserves towards less readily available soil organic matter pools; and other physical, chemical and biological changes affecting decomposition rates.

 

AGE RELATED DECLINE IN FOREST PRODUCTIVITY
(Publications 31, 34, 36, 43, 50, 94, 95)

 

This research has aimed to explain one of the most commonly observed and commercially important patterns in forest ecology: that forest productivity reaches a peak early in stand development and then gradually declines. In even-aged forests aboveground net primary production (ANPP) reaches a maximum in young stands and decreases as stands mature by as much as 76%, with an average reduction of 34%, according to 13 published studies of forest age-sequences in (36). The rate and cause(s) of the decline are of importance in terrestrial ecology, forest management and global carbon budget analyses. For instance, if the decline is sharp, forest managers will harvest forests on short rotations, leading to more frequent site disturbance. Age-related NPP decline is relevant to global carbon budget modelling because forests account for nearly 80% of terrestrial NPP. We investigated causes for the decline in young eucalypt plantations and chrono-sequences consisting of even-aged lodgepole pine stands of contrasting ages. We investigated 3 current hypotheses for the decline: (1) an altered balance between photosynthetic and respiring tissue, (2) decreasing soil nutrient availability and (3) increasing hydraulic limitation as trees grow in height. Results from eucalypt plantations supported Hypothesis (3). Results from chrono-sequences supported a combination of Hypotheses (2) and (3). Until 10 years ago, it was generally accepted that the decline was due to increased respiration costs as trees mature, combined with stable or decreasing canopy photosynthesis, leading to an altered balance between photosynthetic and respiring tissue (Hypothesis 1). However that hypothesis has now been largely discredited.

 

We incorporated a new model of carbon (C) allocation into G'DAY. The new model assumes that C generated from NPP is allocated to stem and foliage so that water transport through sapwood vessels is sufficient to meet evaporative demand on the canopy, without causing debilitating cavitation of xylem vessels. Through this mechanism, hydraulic limitations can constrain leaf area development and hence NPP, even on sites that experience little soil water stress. The above model of hydraulics predicts that the sapwood area to leaf area ratio should increase with increasing tree height. We have shown that this prediction is supported for several eucalypt species, but not for native stands of Eucalyptus delegatensis (50). One of our past Honours students, Karel Mokany, investigated why E. delegatensis is different by measuring leaf and sapwood area and stem hydraulic conductivity, and concluded that the difference occurs because the stem hydraulic conductivity of E. delegatensis increases with increasing tree height (50).

 

 

Our current work on the above topics includes: