Highlights

Instrumentation at the Synlait Farms study site. Image - A Farr

Impact 1: Terrestrial greenhouse gas emissions and removals are understood and quantified so that changes in relation to management strategies, land-use policies and global change can be predicted.

Modelling future forests growth

How fast can different forests grow now and under future climates? These are important questions both in terms of current and future wood supply for the forestry industry, and for assessing the potential of our forests to sequester carbon to mitigate climate change. Past assessments using empirical modelling to provide national growth estimates have limited scope and reliability for future predictions.

In collaboration with the CRIs NIWA and Scion, we customised and refined an Australian physiologically-based approach to modelling wood volume growth and carbon storage for Pinus radiata and indigenous kanuka /manuka stands, two important and diverse forest types in New Zealand. We tested the model against a wide variety of measurements (e.g. short-term daily fluxes of water and carbon exchange through to growth rates assessed and measured over years and decades) and across a wide range of environmental conditions.

This provided a robust and physiologically-based description of P. radiata productivity under current conditions in New Zealand and its likely response to climate change. Stand growth in carbon sequestered ranged from about 10 tC/ha/yr in the fertile, warm, wet western half of the North Island to only 2–4 tC/ha/yr in Central Otago and Canterbury and no growth at all at higher altitudes of the mountains of the South Island. Regional growth patterns were similar for kanuka /manuka, but maximal biomass growth rates were only about 2 tC/ha/yr in the most fertile regions and about 1 tC/ha/yr in regions with more adverse environmental conditions. These differences between the stand types were mainly attributable to carbon losses from self-thinning in moderately young kanuka/manuka stands, and a slowing of growth rates in older stands as trees have to increasingly compete with later-successional species that eventually replace kanuka/manuka stands leading to a loss of the carbon initially stored by those trees.

Temperature is often a limiting factor to growth. With climate warming, stands are likely to increase their growth potential over much of the cooler parts of the South Island and at higher-elevation sites on both islands. In contrast, there are likely to be growth reductions in warmer parts of the country or in drier regions, such as the east coast of both islands, where even moderate warming can intensify existing water limitations. However, even these limitations could be overcome through plant response to increasing atmospheric CO₂ concentration, provided plant responses will be as strong as currently anticipated based on a limited range of experimental observations.
More »

This research was supported by Core funding and MPI-SLMACC funding grants.

Pasture to shrublands, the impact of land-use change on carbon storage

As pastoral grazing is abandoned in many marginal areas of New Zealand, native woody vegetation is reinvading and regenerating. Knowledge about changes in carbon stores can be used to reduce our carbon liability, and while the modelling work described above incorporates such research, some aspects need refining through further research such as in this project.

In a well-watered pastoral site retired from grazing in North Canterbury, we measured the carbon gain each year. In the first six years, above-ground carbon increased by at least 10 times, but it was very labile and fluctuated seasonally.

The major species that regenerates naturally on abandoned pasture are kanuka/manuka (‘shrubland’), and although we know a lot about the mature shrubland on older abandoned lands, there is little information on the early stages of growth, which can be slow. It may take six years until the carbon stored in the trees equals that stored in the grass. But since the growth is exponential, carbon storage quickly increases in the next six years. After four years the amount of carbon in the soil to a depth of 0.3 m has decreased by 9%, mainly due to the decrease in soil bulk density once compaction from stock ceased. Such small changes are difficult to detect and there is considerable spatial variability, but understanding the processes regulating soil carbon dynamics is critical for forecasting future carbon stocks.

These results can now be used for carbon accounting to take into consideration the changes in New Zealand land use. They can also be used to model changes in carbon storage in the early phases of pasture abandonment and forest regeneration, to allocate carbon credits. These models can be employed for national-scale carbon estimates by the Ministry for Primary Industries (MPI) and thus effectively increase the precision of these estimates of carbon storage.
More »

This research was supported by Core funding.

Impact of forest disturbance

We showed that forest disturbance is critical for regulating the balance between forests as long-term sources or sinks of carbon. A new model of carbon sequestration during forest regeneration demonstrated that a suite of disturbances (windthrow, earthquake, beetle outbreak) caused a small net loss of carbon (0.3 tC/ha/yr) in mountain beech over a 30-year measurement period.

This work was supported by MPI-SLMACC.

Nitrous oxide emissions

Nitrous oxide emissions are largely driven by livestock numbers, fertiliser use and farm management, and make up about 14% of our anthropogenic greenhouse gas emissions. Nitrous oxide emissions are of growing importance because of agricultural intensification, but have proven very difficult to estimate because they are highly variable over space, climatic conditions, land management and time. Direct measurement at a national scale is not feasible. New Zealand needs its own specific capacity to estimate nitrous oxide emissions so it can report accurately and develop and deploy technologies to reduce them.

We found that the nitrification inhibitor DCD applied to grazed pastures in the Manawatu can reduce nitrous oxide emissions from urine patches by 55% and also reduced nitrate leaching. Soil temperature is the major regulator for the breakdown of the inhibitor, with the inhibitor lasting longer in soils at lower temperatures. This suggests the effectiveness can be optimised by applying the inhibitor at different rates and frequencies for different seasons.

We improved the basis for estimating nitrous oxide emissions for poultry, and our work led to the adoption of a revised country-specific emissions factor from the default value for New Zealand. This has led to a reduction in emissions liability for the poultry industry by 50%. Through complementary research, we recommended a reduction in the New Zealand specific emission factor for ammonia (an indirect greenhouse gas emitted following the use of urease inhibitors applied with nitrogen fertiliser). This equates to a potential reduction of 47 700 tCO₂-e annually for New Zealand.

Our collaborative research with AgResearch has already led to improved estimates for nitrous oxide emissions from hill country pastoral farming, and is likely to significantly reduce the emissions liability of the sheep and beef sector.
More »

This research was supported by Core funding (including capability), MPI, NZAGRC, PGgRc, Fonterra, DairyNZ, New Zealand Fertiliser Manufacturer’s Research Association and Ballance Agri-nutrients, and is collaborative with Lincoln and Massey Universities.

DNDC model for New Zealand

We used field and laboratory data to develop a local version of the DNDC (denitrification /decomposition) nitrous oxide and methane flux model for grazed pasture under New Zealand’s climate conditions, soils, pasture growth patterns, fluctuations in soil water, grazing regimes, and fertiliser and animal nitrogen inputs, including the effect of uneven spatial distribution of urine deposits. This NZDNDC model, applied at the regional scale (Manawatu), successfully showed where, and to what degree, emissions are a problem.

By running our NZ-DNDC model populated with 20 years’ of climate data over a range of soil types, climatic conditions and farm management practices, we developed ‘look-up’ tables of direct nitrous oxide emission factors for New Zealand. We linked these calculations to a framework allowing us to estimate emissions at large spatial scales. Because this approach incorporates variability associated with soil and climate, it will improve estimates of agricultural nitrous oxide emissions at regional to national scales, and enable us to forecast the impacts of land-use-change scenarios.

This research was supported by Core funding, MPI and NZAGRC.

Impact 2: Strategies for land use and asset management increase carbon storage, mitigate greenhouse gas emissions, and balance environmental, economic and social benefits.

Partnering with the dairy industry

We initiated a 5-year project in partnership with Synlait to develop more precise techniques to measure emissions of agricultural greenhouse gases at paddock scales. Synlait is allowing us to use one of their Canterbury dairy farms and will provide relevant farm management records. We have installed and tested field instrumentation for simultaneous real-time measurements of the exchange of carbon dioxide, methane and nitrous oxide over an irrigated paddock and an adjacent non-irrigated site. Our measurements will start in spring and continue for the next five years.

This research is supported by Core funding and Synlait.

Methane mitigation

There is currently no technology available to manage methane emissions from the dairy-farm effluent ponds where energy capture is not an economic option. In a series of in vitro studies, we identified a suitable soil methanotroph (bacterium that ‘eats’ methane) population able to consume the high emissions typical of these effluent ponds. We built a prototype methane biofilter by incorporating the most active and resilient methanotroph population in a volcanic soil matrix. This biofilter has been consuming >95% of the methane produced by an effluent pond on the Massey University dairy farm, with minimum maintenance for the last two years.

A second prototype, in which the filter is incorporated in the effluent pond cover, has also performed efficiently and does away with the need for the costly plumbing required by the first prototype. This research has brought us much closer to designing a practical, low-cost biofilter. It shows that, by incorporating active methanotrophs in a floating artificial island made from recycled plastic bottles and supporting wetland plants, we could not only capture methane but also remove nutrients to produce harvestable biomass.

This work is supported by MPI-SLMACC and is collaborative with colleagues from Massey University; University of Victoria, Canada; University of Western Sydney.

High country carbon

We know very little about the size of ecosystem carbon stocks and carbon fluxes associated with low-intensity pastoral grazing in New Zealand non-forested high country lands. Landowners and managers have few options for mitigating and offsetting greenhouse gas emissions (exotic afforestation may be inappropriate or of limited viability). In a collaborative project, we assessed whether retiring land from grazing increases carbon sequestration. Results show that retiring land from grazing has negligible effect overall on carbon stocks, even after at least a decade. The largest changes occurred in the smallest pools of carbon (i.e. plant mass and litter). The extensive grazing regimes produced smaller differences in ecosystem carbon storage than the variation that occurs naturally across different sites. These results contrast with those from other agricultural lands (cropland, dairy land) where ongoing losses of soil carbon are associated with long-term land-use intensification.

This project was supported by the Sustainable Farming Fund and was collaborative with high country agricultural landowners, land managers and industry groups, local government and NGOs.

Carbon in cropping soils

Tillage of cropland soils results in losses of soil carbon, but the significance is highly dependent on the degree of disturbance. Our measurements at the Millennium Tillage Trial near Lincoln in collaboration with Plant & Food Research have shown that the degree of carbon loss following tillage of barley is dependent on the fraction of labile carbon in the soil. We used state-of-the-art techniques employing stable isotopes to partition carbon loss from soil and roots, without disrupting the system, to develop a new way to estimate losses of labile carbon. We compared this approach with more traditional, labour-intensive methods that use physical fractionation and found good agreement. Continuous tillage at our experimental site over a decade has resulted in a loss of labile carbon of 1%, compared with a loss of 0.4% in adjacent sites with no tillage.

This research was supported by Core funding.

Pasture quality from space

We are combining observational data and process-based insights from our research and databases in advanced models to improve estimates of greenhouse gas emissions integrated from small plots to catchments to the national scale – as has been described in preceding highlights.

We are also developing models to estimate pasture quality using remote sensing data from satellites in space. Pasture quality has an influence on methane and nitrous oxide emissions, and the current values used in the national greenhouse gas inventory are crude and inadequate. We are undertaking field sampling (‘ground truthing’) across various farm types and regions, and testing the capability of remote sensing to predict pasture quality for future monitoring. The sampling campaign has led to much improved spatial pasture-quality estimates for use in the national greenhouse gas inventory. The remote sensing research could improve cost-effectiveness in the future.

This research is collaborative with MPI, and was collaborative with AgResearch and On-Farm Research.