GCFF Session 1 : Maximise the benefits from the existing forest resource
Presentations: GCFF conference 2019_Session 1
New Zealand forest researchers are lucky to have the legacy of numerous large-plot trials, planted by earlier researchers and now reaching harvesting age. Scion has undertaken end of rotation assessments in some 27 of these trials, and a vast amount of data has been generated as a result.
For example, work on stand density has:
The trials have also been a valuable testing ground for new technologies such as laser scanning, tools for measuring internal wood properties, and UAV-mounted LiDAR.
SEGMOD – a tool to determine the costs and benefits of segregating logs at various points in the supply chain – has also been developed and is freely available to forest harvest levy payers. Case studies using SEGMOD have shown that returns to both growers and processors can be increased by up to 10% through better segregation practices.
Trials to assess the cost-benefit of applying mid-rotation fertiliser to radiata pine have produced promising early results. The work is based on understanding plant nutrient balances throughout the rotation, and how these vary depending on site and the specific crop.
Projected end-of-rotation results from nine trial sites across New Zealand indicate that site-specific mid rotation fertiliser treatments will result in:
Research into the potential for aerially applied liquid foliar fertiliser is also underway. Foliar fertiliser is widely used in other sectors, and has the potential to improve nutrient use efficiency and hence environmental outcomes.
Foliar application by UAV, especially in small woodlots, is believed to have very good potential.
Final stand density is known to be an important determinant of crop volume and value. Scion has developed a model which predicts optimum final crop density for forests under pruned and structural (unpruned) regimes in any part of New Zealand. The model is freely available for forest levy payers.
Research has also revealed the economics of structural regimes in relation to stand desity in more detail than was known before. Growers now have the tools to fine-tune their final crop stocking.
Scientists continue to test a range of remote sensing technologies in the quest to optimise forest data collection. From satellites to aircraft including UAVs, and hand-held tools, the challenge is to generate data at the required resolution in a timely way and at a cost-effective price.
Looking to the future, predictions made include:
The dynamics of a forest depend on interactions between genetics, environment and silviculture.
Machine-learning tools to model these interactions are being developed, and can investigate how changing a particular variable can influence overall stand productivity.
Data from Kaingaroa Forest has been used – some 2.78 million observations on 62 variables have been input into the ‘Catboost’ model, resulting in over 170 million values to be handled. The model has enabled a ranking of how different elements – for example seedlot, spring temperature, elevation, leaf area index – affect productivity.
Results show the potential of machine learning models to interpret phenotypic variability across a forest. Selection of specific seedlots for given areas of the forest to optimise productivity is one example of the many possible benefits for growers.
Techniques using LiDAR outputs have been developed to correlate individual tree crown metrics with other physical metrics of the tree, such as height, total stem volume, and diameter at breast height (DBH). Large data sets have successfully been screened to select individual trees of interest.
In addition, UAV-based data is enabling point clouds of individual trees to be created which describe stem metrics as sweep, lean, taper and branchiness as well as stem diameter and height.
Ultimately scientists will be able to ‘fuse’ above and below canopy data, enabling trees to be measured and described by remote sensing technologies. The benefit to forest managers will come in various ways, not least in potential savings in forest inventory costs.
By combining multiple computer models describing different aspects of tree growth, a virtual tree model has been created, which can ‘grow’ radiata pine internally up to age of 7-8 years.
The model produces high-resolution 3-D maps of internal wood properties, and simulates key internal transport processes such as the movement of water and carbon dioxide.
Ultimately the aim is to produce a tool that will enable growers to make the link between the quality of their product and the growing environment. The hope is that they will then be able to adapt management factors to improve timber quality.
A new series of long-term, large-plot trials, the ‘Accelerator Trials’ have been established at six sites between 2015 and 2018. The trials provide a facility for testing new science, and are already being used to trial some novel establishment and early intervention practices.
The aim is to ‘stack’ productivity gains from a range of different domains and occur at different times during the rotation.
The Accelerator trials will be part of research to understand soil properties and their interaction with tree nutrient balances. Key to this work is the NuBalM model developed by Scion and which has various applications – for example:
Stand density, carrying capacity and stand uniformity are all known to affect overall productivity. The Accelerator trials will enable research into optimising stand productivity through manipulation of these factors for many years to come.
For example, early research into the effect of two different ground preparation techniques on growth of trees in different families has already yielded some interesting and significant results.
Trees don’t grow in isolation and microbial activity in soil plays a huge role in helping tree productivity. Microbes can increase stress tolerance, produce hormones that alter growth and keep resources such as nitrogen available for the trees to use.
Research has shown, for example, that the microbe ACC deaminase can improve drought tolerance, but importantly there is an interaction with tree genotype as well, so different families will respond to differing degrees to the presence of the microbe.
The whole area of soil microbes within forests is very new, with many unanswered questions. We are already moving from measuring microbes to actively managing their activity, however.
The rate at which soil microbes and soil animals release mineral nutrients from organic matter is a key limitation to nutrient availability for trees. Research has shown that a compound called biuret can be applied to young radiata pine and increase nutrient uptake and hence growth. The indication is that biuret helps soil microbes deliver more nutrients to plants.
In addition to better growth, biuret has been shown to increase resistance to dothistroma in one Accelerator trial in South Kaingaroa Forest.
Scion is now supporting efforts to make biuret commercially available In New Zealand.
In summary – the Accelerator Trials are considered a key legacy of the GCFF programme.
Work continues to reduce the risk of soil erosion and debris flows on steeplands following harvest. Research includes:
Research continues into water dynamics under radiata pine; also the movement and uptake of fertiliser and how this can vary according to site management. Results indicate that some genotypes are better at utilising water than others.
Also, having additional organic matter on a new planting site, such as windrows, will reduce nitrogen loss through leaching. This suggests that there are management interventions to consider in areas where the risk of nutrient leaching is a concern.
One project monitored water quality following mid-rotation fertiliser application, and compared the relative impacts of foliar fertiliser and granular fertiliser. Results showed that any impacts on water quality were short-term, and mainly confined to the day of application.
Foliar fertiliser showed minimal impacts on water quality, again emphasising that there are ways managers can minimise the environmental risks associated with their activities.
Scion is part of a global network investigating the impact of harvest residue removal and fertiliser amendment on long-term sustainability.
Conclusions of the research to date include:
In terms of impacts of organic matter removal on soil microbial communities, bacterial communities show high resilience, but fungal communities are much more seriously affected.
In terms of impacts of fertiliser addition on soil microbial communities, the conclusion reached is that adding fertiliser reduces microbial diversity over the whole rotation, especially on sites with low soil fertility.
The Forest Investment Framework (FIF) is a tool that combines spatial, productivity, economic and environmental data, and can assist in quantifying non-timber values of forests to assist with decision making.
Benefits of valuing ecosystem services